Among the ideas floated here, I'd like to add to the city capturing and city razing effects shpould be changed for Civ V.
(Note:this has been stated elsewhere in different form by others)
1. When razing a city of population more than, say, three, you should have the ability under Slavery to make a worker out of each population point, OR add them to a city.
2. When capturing a city it is always a bummer that EVERY building(some few exceptions) is completely destroyed. Taking a city would be much more fruitful if it didn't mean having to build it again from the ground up. My suggestion is to have a percentage chance of destruction for each building type. For instance, walls in some cases would NOT be completely destroyed, and markets and/or libraries probably would be.
But let's take it a step further. If a civilization is in the medieval era and three out of seven cities are sacked, surely their civ should fall back a good ways in developemnt on account of stored knowledge in libraries and monastaries being lost. They should actually LOSE techs. This is a small detail that can be easy to create and would add a realism to gameplay. Likewise, sacking a city should give the chance(percentile calculated) of the invading civ to gain a tech from the capture. Not always, but occassionally, since every civs knowledge base is in their cities, right? And likewise, the size and culture of a city would help to determine the negatives and positives on each side for what is lost and what is gained.
Naturally, this would be lessened a good deal once Paper came along, and much more so as the game advanced, to the point where, in the computer age, it becomes much less probable altogether--but not impossible.
3. Rather than getting too into resource utilization, why not feature certain techs that you need a resource to be able to discover--like iron-working. This would make it impossible for a land-locked nation to discover Seafaring, for instance. Ways that civs are kept backward should be exploited so as to encourage meaningful resource trading to begin with. Likewise, if we carry the religion idea a little further, why not feature some religious techs that are automatically given to those civs that share that religion? This would be some boost to the religion game, as well as a dicey way to suggest that overall religious conversion has it's pros and cons. I like the way religious blocks develope in BTS, but taking it that step forward is a real plus--much like how with the UN you are able to introduce civics to all civs, only in this case it is automatic.
4. Last thing--having some thing integral like the Revolution concept. Really playing with city-state theme, and only being able to really tie together tightly your civ near nationalism. This naturally gives religion thast much greater an impact, but other factors can be introduced that can help keep your city in line. Remeber how in Civ 2 your cities would be demanding certain buildings? Along those lines, only with the culture aspect of it, these cities can gain their own identity! Which makes where you build what wonder all the more critical to your overall empire.
Just a few thoughts.