Yes, I see your point, and always have done. It answers that particular part of the OP: “or has adapted thru evolution the least”.
But it gets in the way of analysis if you claim that nothing extant can be primitive. It’s just a hijack of the word “primitive”, which in this particular case means of or relating to an earliest state, or being little changed from an early ancestral type.
No it’s not. As I and others pointed out, the question actually shows a misunderstanding of what evolution is all about. If you said “little changed in physical appearance”, then that would be accurate. But there is much more to evolution than just superficial physical appearance.
As for my post a abut monotremes, I probably used the word “precursor” incorrectly. I meant to say that it has not been deomonstrated that monotremes, as a group, date back further in the evolutionary timetable than do marsupials or placentals (as per the thread I referenced).
But the interesting thing about our extant monotremes is that they are highly specialized organisms that aside from egg-laying bear little external resemblence to ancestral mammals. The playtypus is a highly specialized aquatic feeder with unique electrical sensors and poison spurs. The two kinds of echidnas have highly derived spines, and a whole host of highly derived specializations for digging and insect feeding.
Obviously, the earliest ancestors of mammals layed eggs. But exactly when that happened is a mystery. We really don’t know how all those mammal-like reptiles of the Permian reproduced. But if it is true that egg-laying is secondarily evolved in monotremes, then it isn’t a primitive condition, and therefore we couldn’t say that the monotremes more closely resemble ancestral mammals, even if we exclusively went by reproduction. That would be like claiming dolphins are more primitive than shrews because they resemble fish. But their marine adaptations are secondarily evolved, not primitive.
This is using primitive in a technical sense, meaning “ancestral”. So, I’d say we can exclude monotremes. Then we can either look at opossums or dasyurid australian marsupials like the kowari (here’s a kowari link: http://home.mira.net/~areadman/kowari.htm), or insectivores like shrews. Shrews look remarkably similar to the mammals that were around with the dinosaurs 100 million years ago.
Also note that the earliest true mammals were small shrew-like creatures, but the ancestors of those true mammals were much larger. One of the ironies of paleontology is that the Permian extinction cleared away the large mammal-like reptiles and made room for the smaller, faster dinosaurs. Then 150 million years later the Cretaceous extinction cleared away the dinosaurs to make room for the smaller faster true mammals.
And by the way, how did a desmostylid get all the way to Sydney? I thought you guys were exclusively North American.
I forgot to add:
Asking “which extant species is more evolved” in biology is akin to asking, in physics, “which celestial body is moving the fastest”.
Okay, John Mace, I see that you and Lemur866 interpret the word “primitive” as meaning “ancestral”.
Lemur866: North Pacific, not North American. 
That’s because, in evolutionary biology, the word “primitive” means “ancestral”.
IMHO, the words “primitive” and “advanced” should not be used in evolutionary biology precisely because of the sort of confusion they produce. Better to use the words “ancestral” and “derived”.
But unless you have an unusual definition of “North” over there in Sydney, I’m still confused.
Ah, when I saw you register last year I slapped my forehead and thought: “That username should have been mine!” You all have the freakiest teeth, they look more like a collection of tubeworms than a mammal tooth. And there was somebody else running around a while ago with the name “Sea Sloth”. There really was such a critter recently discovered, a fossil marine kelp-eating giant sloth from south america, dubbed Thallasonectes I believe. Craaaazy.
The question in the OP would probably be better phrased thus:
The ancestral mammal had several traits which are considered plesiomorphic (a fancy term for “ancestral”), with respect to the rest of Mammalia. All other mammals must necessarily have a number of derived characters. Given that, which extant group most closely resembles (in the sense of possessing fewer derived characters, not necessarily in the sense of “looks like”) the ancestral stock?
[Stupid cursor…I hit “enter”, and thought I was going to another line…apparently, I was actually submitting my reply!]
Anyway, to continue:
The problem with answering such a question has already been explored in previous posts. Even though, for example, shrews “look” primitive, the fact that they are placental mammals makes them quite derived, despite their outward appearance. Even marsupials and monotremes (despite both being considered “more primitive” than placentals) have numerous derived characters relative to basal mammals. And, of course, those derived characters are the direct result of selection and adaptation. So, it’s really not possible to say “which mammal is least derived”, especially since “least” is not quantifiable in this case.
I just reread the OP, and he’s actually not asking which animal closely resmbles the acestral mammal (whatever the hell that is), but which extant species has remamined physically “unchaged” the longest. Again, difficult to quntify, but it is a different question (even different from the title).
Although the OP says “hundreds of thousands” of yrs, let’s say he really meant to say, just the longest time, period.
So, again, I think it would probably be something like a shrew, since most mammals we see today don’t look anything like their ancestors from 60M yrs ago.
Well, in the sense of retaining the greatest number of ancestral (theraspid) characters, I would have to go with monotremes.
From here
However, as has been previously pointed out, extant monotremes have a large number of highly derived characters. Therefore they couldn’t be considered to have the fewest derived characters, even if they have the most ancestral ones.
I would go with shrews among the placentals, and didelphid opossums (or possibly rat oppossums, or caenolestids), among the marsupials, as the most conservative groups relative to presumed ancestral types. As placentals and marsupials are of similar antiquity, it is not really possible to say which is more derived.
However, as Stephen Jay Gould delves into in one of his essays, the echidna is not “primitive” but adapted in non-Therian ways – its brain is remarkably advanced, albeit with different neurology than the rest of us mammals have. In short, it has adapted to be very efficient as an echidna.
The only extant mammal family surviving from the Cretaceous on is the Didelphidae, the opossums, with genus Didelphus dating, IIRC, from the late Eocene or early Oligocene. It would probably be the winner in a “living fossil mammal” competition.
Among the placental mammals, most of the lipotyphlans, the “standard” insectivores, are adapted quite efficiently for their lifestyles, but in ways that have not changed greatly from their ancestors at the earliest of therian-mammal times. By a minimal-degree-of-change format, it’d be a tossup between them and the didelphids.
The genus of large placental mammal that’s been around the longest would have to be the Diceratherium, which was the common rhinoceros of the Miocene and survives today as the Sumatran, or Asiatic two-horned, rhinoceros.
We can pretty much discount any terrestrial Australian species. Evidence suggests that all Australian species have descended from arboreal South American species. As such they can’t be ancestral. The opposums would be a more likely candidate.
As far as that question has any meaning probably the pangolins. There are fossil pangolins from the Eocene which are almost indistinguishable form the modern species.
Not quite. A cretaceous platypus skull of the family Ornithorhynchidae has been found. So there were platypus swimming around amongst the dinosaurs, as well as opossums running through the trees.
So is there any general feeling about where pangolins fit in? They don’t seem to have any close relatives, do they?
So let’s create a rigourous version of the spirit of the OP’s question. How about “which currently existing species could be transported back the furthest in time and still be cross-fertile with an existing species of that time?”
Is that an acceptable definition of “least evolved” for the sake of this thread?
It wouldn’t fit Little Nemo’s definition, but I offer Homo sapiens: very little evolution since the formation of this species some 150,000 years ago. Of course, the newer the species, the less it will probably have evolved (while being that species).
Good point, amarone. I bet Terminus Est wishes he’d thought of that: “In which case, Homo sapiens sapiens would qualify, having emerged only about 120,000 years ago.” 
I can’t see how any answer is possible with your version Little Nemo. It could only possibly be answered by actually breeding the animals. Many animals today look very dissimilar and are interfertile (false killer whales and dolphins, llamas and camels). Conversely many highly similar species are totally infertile (cougars and jaguars). Hard to make decisions on fertility based on fossils.
Even if we could do this, would it tell us anything about rates or amounts of evolution?
Consider that camels and llamas diverged about 20 million years ago. If we assume that they were even half as interfertile with all common ancestors as they are with each other then we could take either back 30 million years and obtain hybrids.
But dogs and foxes only diverged 10 million years ago and are totally sterile today.
Many animals can easily top that. Grey wolves completely indistinguishable from modern forms are known from 1 million years ago.
Not really, since the cross-fertility is generally something used to define species in the first place. And if we are talking species, then the question would essentially be “which extant species has the longest fossil record?” Unfortunately, there doesn’t seem to be a lot of information about the fossil records of extant species available.
Cladistically, Pangolins are in the group Pholidota, which is a sister group to Epitheria (which contains all eutherians except Pholidota, of course, and Xenarthra [anteaters, armadillos, & sloths]). Their closest relatives (based on both morphological and genetic data) appear to be members of Carnivora.
Completely unanswerable. The best way to look at this is: “What animal most closely resembles (physically) it’s fossilized ancestors the fartherst back in time?”
I’m pretty sure that’s what the OP intended. Speaing of which, where is Aslan2? Probably wanted a simple answer, saw all the bickering, and just decided to stay out of the fray.