And the problem with Civ V and VI is not that they are bad games.
The problem is that Civ IV is way better.
That sums up my feelings on Civ V. If I'd never played a previous Civ, I'd probably be pretty happy with it. But I have, and Civ IV and Civ III are both much better, IMO. I recently started both a Civ III and a Civ V game... the Civ III game heated up pretty quickly, soon I had a half dozen cities, then more, then an ancient war, then Montezuma declared war on me. In the Civ V game, the pace of expansion was glacial by comparison (and I still hit the happiness penalty), although I did finally manage to lose a city for the first time ever by completely neglecting my military and sending literally all my combat units out exploring for 2500 years. And whereas losing one city in Civ V is a big deal because you have so few (and if you don't 100% neglect your military, it's almost impossible to lose one), in my Civ III game I lost a few cities, but had enough that it wasn't game over, and was able to fight back, retake them, make peace, have the Aztecs invade again, lose another city, etc. - it was a lot more fun.
And Civ IV trends closer to Civ III in that regard. Civ IV definitely favors defence more than Civ III does, and the overall number of cities is in between Civ III and Civ V, but the overall pacing and scale is much better than V, and it adds some nice additional options that III doesn't have. And it always amazes me how both III and IV have a noticeably better AI than V. Not that III and IV's AI is good necessarily (at least without mods), but it didn't take me 5 years to lose a city in them, either. That's something I hope they will have fixed in VI, but I'll wait and see the reports on it first.
Eurekas are another thing that I want to wait and see on. They sound potentially interesting, but the same could be said of city states in Civ V, and IMO they were a bust, especially in Vanilla. I'm wondering if a more graduated approach may have worked better. For example, instead of 50% off for building 3 farms or whatever, how about 10% off per farm up to 5 farms? Or even scale it a bit - 10% for one, 20% for three, 30% for five, 40% for eight, 50% for twelve? Point being that a 50% discount is big, and an arbitrary line in the sand has been drawn. That both can skew strategy - it may be worth doing something that makes no sense otherwise - and imposes a significant penalty for being just short. Whereas with a more graduated approach, you're still giving bonuses in areas to civs who do actions that are related to them (the stated raison d'être), but the bonuses you receive would evolve more naturally from your surroundings and playstyle than "here's a random spearman for a Eureka bonus".
Another idea would be to have in part the specific-Eureka-bonus-per-tech already shown, but also in part general skills in a certain area playing a part. For example, each tech may have a category - agricultural for instance - and performing agricultural actions - building farms and granaries - would increase your agricultural knowledge. You could get up to 25% off any agricultural tech for having maximum agricultural actions (and a lesser bonus for being somewhat agricultural), and perhaps another 25% for actions that vary per-tech, as currently Eureka bonuses do. That would provide some longer-term continuity in bonuses. You could even tie this in with production perhaps; a civ that constructs lots of large regular buildings like Aqueducts and Colosseums may be quicker at building Wonders as well, since they already have the expertise for large-scale construction (although I do like what they're already doing with more wonders being map-location-dependent; they did this with one wonder in Civ III, and there are resource bonuses for some wonders in Civ IV, but at least in concept I like that it's being expanded in Civ VI).