For example, if I wanted to patent a mechanical toothbrush powered by the rowing of 6000 men, could I do so? Since rowing machines and toothbrushes are already patented, can I still patent my invention?
Also, it’s apparently possible to patent marketing techniques. If I wanted to patent the marketing technique of standing on a street corner yelling “Extra, extra! Read all about it!”, could I do so? And if so, how would I enforce such a thing? Would I be entitled to charge people that used that marketing technique without my permission?
1] It has to be new (‘novelty’ aspect), so you can forget about claiming “Extra, extra! Read all about it”.
2] It has to be surprising (‘inventive’ aspect).
3] It has to be better than existing things.
I think the third aspect would prevent your new toothbrush.
Obviously,  is a tricky one and leads to lots of discussion.
A good example:
Engine-less gliders are known (and say even patented). Now someone comes up with the idea of having an engine with a propeller added to such a machine. Movel, inventive, improvement, and a patent would be granted.
The next inventor realises that fixing such an engine/propeller on each wing would be even better. Sufficiently different (innovative) from the original one to be patentable.
Now I try to patent one with two engines/propellers on each wing. Most likely a patent would not be grnated, as this is a clear extrapolation from the second idea.
The first one: yes and no. IANAPA, but I think you have to be able to demonstrate that you have made the invention if so required. By the way, if someone else has discovered it before you, but did not take a patent or otherwise made it public, you still get your patent. The other guy/girl does not have to pay you money though if he wants to use the idea as he can show he had it before you.
The second one: sure. You can give out licences for usage to cash in, but you can also sell the rights to your patent to someone else or a company. Or it can be traded for other patents.
And if you plan to patent a pertetual motion device, you have to provide the patent office with a working model, not just a description or drawing, that will continue to work for long enough (a couple of years, I think) to convince the authorities that it isn’t a hoax.
The patent office had to make this provision because they used to get so many applications for perpetual motion devices and decided that they had to draw a line.
At least one author mentions this in a book about aviation history: the patent office was getting so many applications for another well-known impossible device (the Flying Machine) that they required working models. A top physicist had proved that powered flight was impossible. Because of this, it took the Wrights about a year of wrangling to get their patent accepted.
I’ve read one article about the “disproof.” The author said that powered flight was judged impossible because of wrong assumptions made about lifting force of wings, and wrong assumptions made about the power output of the engines of the time. In hindsight it appears that the disproof was probably motivated by pseudoscience. There are two kinds of pseudoscience: the one where a scientist is convinced that a false idea is true, and the kind where a scientist is convinced that a true idea is false. If “everyone knows” that flying machines are a crackpot pursuit, the professionals would tend to be biased against them and to attempt disproofs rather than pursuing openminded research.
Note that theory and practice at the USPO are two entirely different things. People patent trivial, obvious, well known things all the time because the USPO has gotten tremendously lazy and doesn’t really do it’s job anymore. The stuff that has gotten patented in the computer field recently is astonishing. I saw a place that just gotten a patent on an event counter for a task queue. I.e., increment the counter when a task is added. Decrement it when it is removed. Not only is this obvious, it has appeared in text books for decades. Hardly patentable. But they have the patent. And are in a position to sue everybody that has written any task based programs. Since fighting such lawsuits cost money, little companies have to bite the bullet and pay a license fee. Large companies work out “phony patent swaps”. (I.e., we won’t sue you over our phony patents if you won’t sue us over ours.)
So as the OP, you bet you sweet bippy you can probably patent any nonsense, and threaten people too boot.
It’s disgusting and the courts ought to do something about it.
Steven Olson patented “A method of swing on a swing is disclosed, in which a user positioned on a standard swing suspended by two chains from a substantially horizontal tree branch induces side to side motion by pulling alternately on one chain and then the other.”