A few notes on a couple of your points, Pere:
-Can't change tax rate is a good change. It allows you to control approval on a planetary basis, where it matters for this game.
I'd heavily dispute your second sentence, tbh. You can't really control morale at all. That's the big reason why LEP is so hated by those playing on bigger map sizes; you're fairly helpless to do anything about it. Oh, you can build buildings for it, which means knocking down other stuff and waiting for ages for it to switch... but tbh, when there's rioting in the street I think most governments can come up with something better than 'we're building a new leisure center, it'll be ready in 6 months'. As I said above, I don't care much about the lack of tax rates either... but really, the game could use an on-the-fly morale improvement ability where the player can trade something to keep the people happy. Tax rates fulfill that role fairly well in most games. If SD can come up with something better, then I'd have no problem with it... but atm, the role is just empty, and that's not a good thing.
In most Civ games prior to 5, happiness could be controlled on-the-fly using tax rates, garrisoning, AND specialists. That gave a player lots of cool options; there were multiple strategies to dealing with morale issues. GC3 doesn't really have any on-the-fly options right now. You build morale buildings, and occasionally you get a lucky Artifact find or tech giving a universal boost. Those are long-term options, requiring considerable investment or acting as eternal upgrades. There's no quick sacrifice that can be made elsewhere to get a boost to a specific planet. If you suddenly get an event that nukes morale for 50 turns, then you cannot do anything about it short of tearing down a bunch of factories across the whole empire, building the approval buildings, and then swapping them back later.
-The economy is not at all like Civ V. It is all based on population, then buildings are multipliers. Whatever your qualms with CGIII economy, it does not remotely resemble Civ V.
Much of the Civ 5 economy is pop-based with multipliers too. Research generation, for example, is entirely based on pop in Civ 5 (which was a very bad change for lots of reasons, in particular because food, industry and wealth remained yield-based). I do see where the OP is coming from on this one too, though really a properly designed pop-based model isn't problematic (and GC3's fundamental model is much, much better designed than Civ 5's. Just hands-down. Food and growth could do with some work, but in most other areas it's a superior design, having stuck with the pop-prod model rather than going for the bastardized schizophrenic mess Civ 5 ended up with).
I think the GC2 model was a really nice attempt at something a bit different and a bit new... but was also painfully flawed; there were silly 'optimum solutions' for almost every build. I think the GC3 model has better potential to avoid that kind of thing, once they're figured out the right numbers to put into it. That's just balance, and so not something I'm super-worried about really; you can endlessly tweak econ values. The relative values of Money, Research and Industry are currently really imbalanced, though, and need some attention to bring them back into line.
-Large empire affects AI and humans the same, and in fact, humans will usually have larger empires and so be effected more severly
Ah, but the human won't stop expanding his empire because of it. The AI will. I've noticed many times that the AI will not colonize past around 80 planets with morale-based LEP, since it's building from a script and so doesn't build extra morale buildings to counter it. The relatively minor penalty for having no morale is something the human wisely ignores, but the AI freaks out about it.
And that's without even touching on how it doesn't actually act to slow down empire growth or extent, or how it's literally incompatible with the scale of the game on larger maps, or the inherent conceptual design flaw of using a colony-scale penalty on an empire-scale problem (which you just shouldn't do - if I can only counter a penalty acquired at empire level by redesigning all my individual colonies, then it's a bad, micro-intensive design). But I've outlined that in exhaustive detail elsewhere on the forum. There's little real excuse for LEP in it's current state and the sooner it's replaced with something that actually does what it's supposed to do, the better.
Not true. with high diplomacy skill, trade routes, and other bonuses, ideological differences are not that hard to overcome.
The OP is just noticing a symptom of the issue, and you're just using another symptom to counter it. It doesn't really clear up the problem. Which is a shame, because there's actually clearly a hell of a lot of work gone into the diplomacy module which is being spoiled by a few silly bugs and an errant design decision or two.
Really, the problem with diplo is that most effects are huge compared to the relatively small relations scale. You have 20 possible positions on the relations scale - -10 to 10. Yet several of the relations modifiers give +3 or -3 PER TURN to this score, and they stack. It's entirely possible to go from best chums to absolute loathing in the space of a couple of turns, based on some pretty minor effects (like ideology, or killing a pirate for them... or your scout ship being in their territory for five minutes). There's a diminishing return in play, but it can't really counter the modifiers acting on the undersized relations scale.
Ideology, for example, has an effect of -2 per turn, increased for some personality types. That means it would take just ten weeks of minor ideological differences to turn firm friends into absolute arch-enemies. The diminishing return acts to limit this, but none the less the impact is HUGE, particularly early on when you have no trade routes and little Diplo skill (which is itself vastly overpowered as well). Honestly, I think the game really needs the diplo scale increasing to 100 to -100. It's really quite nuanced and nicely put together, with personalities having lots of effect... but no-one can really see it, since the impact is basically 'they'll hate you in 4 turns instead of 6!'.