1) its not possible to expand to fast if all the worlds are colonized before you get to 20 colonies, therefore it solves the problem of expanding to fast.
You want the whole galaxy to be colonized by turn 18. Riiiiiiiight. I think you're advocating removing at least 2 of the 4 X's with that. Also, this borders on the 'settings other than my preferred ones don't need to be balanced properly' fallacy.
2) You can make the LEP scale however you want, it's currently linear which is meh. If tuned correctly the LEP does this:
-small empires have high happiness and get more production/growth/etc to keep up
-large empires have medium/low happiness early/mid game because they expand fast, but grow very strong late game.
-very large empires that expand to fast suffer from rebellion/striking workforces and stuff like that.
You're almost right. But LEP has no relevance to the RATE of expansion - only the limit. None. Not a sausage. I get the same LEP from 80 planets whether I got them by turn 12 or turn 1200. If anything, it encourages me to get them as early as possible, because why the hell not? I'm gonna hit that LEP cap sooner or later, and if I can hit it first then I will remain ahead of the other empire indefinitely as they cannot breach the LEP cap either. It is a soft cap on total planets. It does not relate to the rate at which they are colonized. This is a small distinction, but the difference in gameplay terms is huge.
The present system encourages colonizing as much stuff as possible, as fast as possible. This is regardless of whether you've carefully min/maxed your race for rapid expansion or not - Marigoldran has taken this to the extreme by optimizing his race as much as possible for expanding, but the strategy he's following (expand really, really fast) is the correct one to use regardless of your race picks.
LEP should be abolished, because it's a red herring - it doesn't address the real problem, and yet people will initially assume it somehow does and point to it with the same flawed logic you've used here. Unless we have a rate control rather than a cap control, colony spam will always beat any other strategy in the early game.
A alternative/combination is to make maintenance cost prohibitive for worlds with low population meaning you go bankrupt if you expand to fast. but i dont like that approach because it incentivizes you to not spent your starting money. But really low happiness achieves the same without complicating stuff, you can easily make a low happiness world a burden by tweaking some numbers around.
Well, first up doing the same thing via happiness is actually more complicated, since you're dragging the morale system in for no real reason. Secondly, it doesn't actually work that way, for the reasons outlined above (cap on total vs cap on rate). Oh, and thirdly, I don't think it really does incentivize not spending your starting cash - investing it in cash generating infrastructure is smarter than just burning it down as a runway. It encourages 'tall' empires over 'wide' empires, which is a good thing. Currently, Marigoldran's game is going quite swimmingly despite the fact he's only developed 1 tile per world.
In the end there will always be a cap on how many worlds you can sustain, and things to exploit and what not. Might as well keep it simple and tie it to one mechanic rather then spread i out to much.
Why should there be a cap on how many worlds you can have?
That's not a dumb question - if we have a game where some maps have only 2 or 3 planets per player, while others have 100 per player, then on the vast majority of maps there won't be a cap at all. There simply won't be enough planets in the galaxy for a cap to come into play. The huge scalability of the galaxy actually requires that we DON'T cap the number of planets that can be colonized.
Currently, the maximum number of planets the AI can support before LEP kills it is about 80. This won't even come into play in medium maps and below. Why should those of us playing on bigger maps be gimped by it, then? Who, exactly, is LEP meant to be balanced for? In fact, what is LEP for? The player can just ignore it. The AI won't ignore it, and so is therefore gimped. It doesn't really have that much impact until you have 30+ planets, which the majority of players will apparently never have. So get rid of it. It's useless. It's only affecting players on maps where it shouldn't be set so low anyway. Replace it entirely with something that is actually fit for purpose and which scales automatically - so something which actually slows the rate of growth, rather than doing nothing to rate and just impacting maximum size.