-Population gives you credit for each 1,000M pop, whereas research, economy, and military give you credit for each 10, 10, and 20, respectively, with regards to score. The upshot of this is that there is no insane population game that will also give a respectable score. Trust me; I've tried. Fast growth does benefit score, but there's only so much growth you can get, and only to certain population levels.
-Immense scores approximately 10.55% more than Gigantic, all other things being equal. However, all other things are not equal. For one, TA Gigantic/Immense appears to be limited to approximately 500-550 planets, as opposed to DA Gigantic's 700-750+. Secondly, the extra distance means almost regardless of what approach you take, even using the Swindle method, it will take longer to 1) obtain all the planets 2) build them up and 3) for your ships to reach the MSBA. Notice I say almost, though.
-TA does have some beneficial aspects to it, mostly coming from the new tech trees, which allows one to achieve almost or equal to a DA score (with a LOT of work), thanks to racial bonuses, and the fact that (spoiler alert!) the purchase now reduction from Neutral alignment and two tech trees (nontradeable, sorry chaps) actually works in TA, whereas it has never worked in either DL or DA.
-At some point the approval bonus of a flat 10% that was not depreciated by population from choosing Neutral became a 10% morale bonus (which is depreciated by population)-but it's still useful.
-While it is true that you can achieve 20k BCs per DL AI per turn (aside from the slot 1 AI), up until I shared this knowledge with Mumble my (possibly flawed) understanding was that no one had done so-or even tried, for that matter. Technically speaking, then, we can assume the difference between a DL AI game and a non-DL AI game to be (roughly) 20k BCs per DL AI-but not per turn. It is worth noting, however, that this refilling of the BCs does enable you to get significantly more cash than you otherwise might for early techs, as one of the factors affecting an AI's perceived value of a tech is how much money they have-below a certain threshold, at least.
-SirPleb, since the ZYW fad took over, and perhaps prior, it has been customary to refer to the first game year as Y0. That is, if your score for Dec 22 Y0 was ~600k, you stand a decent chance of cracking 2M. However, if it was, as you said, Dec 22 Y1, then you quite simply don't. Semantics, I know, but clarity of communication is vital.
Finally, I've done some testing in my game at home (no internet connection at present), and I can't get consistent results. Either I'm doing something in my admittedly shortcutted testing to somehow change the base growth (which I wouldn't discount out of hand since I tend to test in Cheat Mode), or there actually is some random function that determines the base growth for a race. Generally this seems to be 4% or 4.5%, but I did see 3.5% once or twice. I've even seen a 4.5% for Custom, which I've never seen anything but 4% for. I haven't done much testing with the Torians as of yet, but I'm fairly well convinced they'll stay at 7% regardless.
My belief is that this is a change starting with 2.x, as I don't remember inconsistent results in 1.8g, but it's hard to know for certain. In any case, 4% base growth still seems to be a decent assumption when we run theoretical population growth numbers.
It's certainly possible that if it IS random you've lucked out and gotten an Altarians with 7%. But, as you say, it really only matters when you're below the 75M growth point.
There is one other thing that I'm still looking into, but it's too early to say anything about it yet.