Can you cite that they do this, or is this your theory?
Er, amend my previous post, there: I meant to say less obscure than “a’a”.
No, no cite, but that’s my belief. Do you think they don’t load up on two-letter words beyond what their usage in speech and writing would call for? Does anyone? I didn’t think that was controversial.
Ok, well I don’t. I’d be a little surprised to find out this is the case.
The basis for inclusion in the Scrabble Players’ Dictionary is that a word has to be found in at least one of five collegiate dictionaries: Funk & Wagnalls, Merriam-Webster’s, American Heritage Dictionary, Webster’s New World Dictionary, or Random House.
The exceptions are (1) the Players Dictionary doesn’t bother with words beyond eight letters long due to their rarity in home play; (2) there was mid-90s purge of swears and ethnic slurs. Both types of words are included in the Official Word List that governs tournament play.
So, if you think ‘hm’ or ‘za’ shouldn’t be words, you need to complain to the compilers of whichever of the five dictionaries decided to include them.
(List of the five dictionaries comes from Stefan Fastis’ “Word Freak.”)
I’ve heard serious players comment that the addition of QI has significantly changed the game. Where as serious players usually exclusively knew the high scoring odd words (XI, XU, JO, QAT, TAJ), ZA and especially QI has allowed novices to garner big scores. Especially in the case of QI where U-less Q words are rare.
A fact that threw a wrench into ESPN’s early efforts to broadcast Scrabble tourneys IIRC. 
Interestingly, this past summer when I was home visiting family, there was a friendly Scrabble game going where folks where trying to help Gramma use the last of her letters so she wouldn’t be stuck with a big point deduction. I suggested “qi”, but nobody else was sure if it was a word, and it certainly wasn’t in the fifty-year-old dictionary we dug out. This would have been after it was added to the tournament dictionary, but I didn’t know it at the time.
The main thing that’s really nice about using the OSPD (aside from it being an agreed-upon standard) is that it’s completely explicit about what is and is not a word… which may sound like an odd comment about a dictionary, but if you look up “dig” in the OSPD it specifically and clearly lists not just “dig” but “digs, digging, digger, diggers, dug” etc, and will tell you whether or not “diggings” is a word. Regular dictionaries can be less explicit about things like that. In addition, regular dictionaries will have prefixes and suffixes in there by themselves confusing you about whether “pre” is a word or not, etc.
I thought it was three out of five or something like that. But the point remains… if “qe” starts being used to mean something, the makers of the OSPD will not consider its impact on gameplay (which would be huge) when deciding whether to include it.
This is backed up at ScrabblePages, for the first edition at least. I’d say calling the OSPD “permissive” is accurate.
Also from that link:
They may not have a policy of “loading up” on useful words, but I’d expect those to be the words that people will request, so that would be the effect.