Buffing a strong situational ascendancy perk which is available from the very start with no preconditions into an extremely strong ascendancy perk with Galactic Paragons, changed to be both extremely powerful early game from the leader capacity, and late game from the max leader level, seems a bad idea to me.
The game has a +2 perk for commanders (Galactic Force Projection), officials (Imperial Prerogative), and scientists (Transcendent Learning). All three are strong, but not every build needs them all.
Also, I beg to differ. It is not "just a fact" that "we" utilize commanders and officials more than scientists. It depends entirely on the playing style of a player in a given game. So if you do so in all your games, I am certainly not going to say that you are doing things wrong because presumably it is either great fun for you or the best thing to do given your playing style, or perhaps it is both, but generalizing to the entire player base is: This is a YOU issue.
If, for instance, you play mostly peacefully and occasionally steamroll an AI, you are highly unlikely to need more commanders than scientists. Sure, you'll probably end up with more fleets than commanders, but that's not an issue as the few times they are called on to fight they'll be doomstacking with one of the fleets that has a commander or you'll be throwing them at enemies that are so much weaker that the absence of a leader has negligible impact.
Officials, you are going to want up to two for Federation and Community, if you've got the former and care about the latter, but since they can work council positions at the same time and the delegate veteran class is awful, this does not necessarily require you to increase your cap. Apart from that, you'll want one for every important sector that isn't predominantly a research sector and, possible, a specialized one for some planetary governors.... How many this adds up to depends to a large degree on how wide direct control you play.
As for TL itself, it is probably rather less situational than you think. The reason I say this is that amongst the few reasons you can think of for picking it, you leave out the main reason that
I pick it most games, because I am a compulsive optimizer.
For my own part, I consider Transcendent Learning to be one of the most attractive ascendancy perks in the game for the two first perk choices, because the experience bonus is huge and will aid throughout the game, and more importantly the +2 scientist cap is great in the early game as your very first perk, increasing your scientist cap from 3 to 5. So you can have 5 scientists out surveying and finding anomalies without a loss of experience or paying a unity upkeep surcharge, or 6 or 7 with only fairly small penalties compared to having more scientists with a cap of 3
That does, of course, have something to do with my settings. I play large or huge galaxies, usually with the default number of opposing empires, which means lots of space to explore and survey for goodies
if you get there first. Somebody playing smaller galaxies or greatly increasing the number of enemy empires to cut down on the room for exploration might well have a very different experience.
