My personal veredict:
What Vainilla Civ V improved over Civ IV:
- The combat. Jesus, the sweet, sweet combat. Killing stacks of death is the best game design decision ever. First time I ever founded myself waging war just for fun, a thing I did not made since the original Civ II
- No roads needed to access special resources. Needs to be copied on further civilization sequels, me thinks
I'm equivocal about this. I prefer the need to connect resources from a 'realism' standpoint, but like transports it was essentially make-work in earlier games that was necessary but pointless in that it didn't add anything to strategy or decision-making; moreover I like the requirement to pay maintenance on roads.
What I do think should be reinstated is that luxuries should only be transportable across a trade network - so a city will have access to any resources in its borders, but other cities will only gain the benefit if connected by trade. This obviously makes no difference for luxuries, but would be important for strategic resources.
- The city - states are a good idea, albeit poorly implemented
I don't have any major problems with the implementation, although in my experience they're (sadly) more passive than in vanilla during wartime, and the more numerous quests now make it a little too easy to gain influence. Only other minor niggles are the road quest, which is still useless (unless you're Austria and then annex the CS you've just earned favour with), and I actually miss the "destroy city state" quests - sure make them rarer, but getting rid of them altogether costs some character. The Patronage opener may now be a little too powerful.
What Gods and Kings have improved over Civ V vainilla:
- Builder style of play is not as boring as in vainilla Civ V. Religions greatly add to this
- AI is vastly improved from vainilla, as well, allowing different gameplays other than "I am surrounded by psycopaths, kill or be killed"
I managed plenty of games without that in vanilla, but diplomacy is certainly more accessible now - it's easier to found those early alliances and to keep them going, though it's in danger of becoming too passive like Civ IV diplomacy and just ticking away accumulating positives so that you can ignore your faithful ally for the rest of the game without worrying about any change in relations. I think it has avoided that trap and that diplomacy generally works better in Civ V than Civ IV, but there are a few missing options like map trading and international trade. Denouncements and DoFs are now a very nice inclusion in the system.
What Civ IV (still) does better than Gods and Kings:
- City placement used to be a relevant factor, and one of the funniest thing of the series, for it forced you to plan quite a lot in advance, but not anymore it seems. Hexes, REX and tile - buying ended up with it, a core gameplay design issue with little chance of being adressed in a expansion
City placement is still very important, it's only the tile output for the city itself that has been standardised. However many tiles you buy you can only work a certain number round your city, and terrain and resource bonuses/penalties apply just as they always did. You still get extra income from river tiles, none (Petra aside) from deserts, boosts if you settle around multiple workable luxuries, strategic resource access, defensive considerations... None of that is in any way "irrelevant".
- Cultural victory and cultural mechanics in general. Waiting for the social policy bucket to fill up is much more boring than being rewarded for building megacities like a champ.
I'd say waiting for megacities to reach the threshold is much more boring, personally.
- Interface. Civ V looks far too cluttered, and numeric values and color circles are less intuitive than ye olde "bread / hammers / commerce" thing, me thinks, but I know that this is a very personal opinion
I can't stand the look of the interface, but then I haven't liked the look of Civ interfaces since Civ II.
- There's little to do during peace time, sans religions. Not quite a lot of decisions to take other than the social policy / belief choosing. I liked how even a single tile improvement could have impact over the course of the game
There's always little to do during peace time in Civ games, other than building (and now you actually have to choose what to build rather than being able to build infinite maintenance-free copies of everything as in Civ IV), and moving the slider every so often if your income dropped into the red.
- Enemy AI regarding diplomacy. Sorry, but this has its roots on core gaming design decisions. The whole Civ V philosophy of "AI playing like humans" and "AI playing to win a boardgame" and "let's ditch the simulator aspect" instead of the more realistic approach of Civ IV makes for a decidely more boring, more predictable game
The AI needs to be better-implemented to play like an AI trying to win rather than like a human trying to win, but I'd say Civ IV's passive modifier-led diplomacy was the more boring and predictable system. I could always guide exactly how a game would play out and, wildcard Monty aside, manipulate the AI civs to be as friendly or as hostile to me as I wanted.
- Variety in gamestyles. This is what I truthly loved about Civ IV, the fact that several different strategies and playstiles worked rather than "the biggest civ wins". You could make trading empires, sprawling empires, small yet sturdy civs, specialist based civs, popullation based ones... each strategy worked on its own way.
Depends on the difficulty level - again my sense is exactly the reverse, that bigger was always better in Civ IV as much as in earlier games, with the exception of the cultural victory condition. That was a fundamental product of the way commerce worked in earlier Civ games - more people in more cities = more commerce = more science. More cities = more people, and with no global constraint like Civ V's happiness system, more cities could always produce more people than fewer. Civ V's global population constraint and base rate of science per pop are specifically targeted at overcoming this historical Civ problem. Yes, if you want to sandbox you can in Civ IV, possibly more so than you can in Civ V, but in terms of selecting "the best" strategies Civ IV was extremely restrictive.
- Much more realistic approach that replicated history mechanics. I loved how it was complex to assimilate a conquered empire, how a bigger bigger did not always meant more dangerous civ or the whole Vatican politics that the apostolic palace allowed. Much of the historic flavour of Civ IV was lost, and quite a lot of the picks of civilization traits and leaders shows a distinct lack of historic understanding from the CivV creators
That's not unfair, but you can hardly accuse Civ IV of historical accuracy in its often seemingly random trait assignments either (and they didn't even get the plural of Mali - Malian, not Malinese - correct), and you had mechanics like whipping slaves to death to boost production that owed more to Hollywood than history.