I don't know how to answer your questions because I don't know what is exactly your purpose with this mod. For exemple, in my own, my purpose is to make more realism and historically accuracy, but without revolutionary change in the gameplay (except if I found something really, really amazing). And get rid of those things which always bug me in civ4 : siege engine destroyed after attack, the whole world always judaist or hinduist, a game ruined by the lack of only one ressource, etc.
But when I made a change, it's with a precise goal. For exemple, terrains. I always think it'a shame there is so many useless terrain, while the map is already a big sea. So, I choose to give some interest for all terrains. It must already be far more interesting to build cities in grassland, BUT, it should be possible to build not so bad cities in second choice area with desert or tundra, and as the game avance in age, new tech make thoses bad tiles better (and maybe with luck, new modern ressources appear not so far of them). It's a difficult change because the bonus should not make thoses terrains better than best temperate tiles. And too much change in a row make the whole thing impossible to balance with impredicable side effects.
For exemple, when I decided to make carry buildable everywhere with a little chance to discover stone, I haven't think the AI will choose them over workshop (because carry can be built before), spam them everywhere even in grassland when a city is in need of hammer, and... spawn stone everywhere in grassland...
So, at your question "should I create some desert improvement ?" I don't know what to answer. I think it's a good idea to give interest to some useless tiles, but, it should have a sense in the whole picture of your mod.
play with the building deletion option - that is part of its goal - those will be removed.
Thanks, I have forgot this option. But I don't think I like it : religions seem able to disappear sometimes, and I really don't want some of my building I have work hard for building, just vanish like that
it is not either realistic, a cathedral don't vanish even if belivers number decrease dramaticaly. In France there is church everywhere while nearly nobody still do the rituals, and they are still part of patrimony, people care to them (more than God lol ).
Civics thats my weak spot, every time i edit those , im not happy with the result.
Maybe it's because there are sooo many. I think there is too many, I don't know what it add to the game to have so numerous choice, and it give an advantage to spiritual leader even though it is already a powerful trait (I think).
For modding civic, I have three way of seeing things :
1- try to keep the original ones, and just balance a bit
2- rewrite all, with a gameplay purpose -> make some for warmongers, some for peacekeeper, etc. Try to match them with real ones but, not so important, it's arbitrary. Maybe "monarchy" can be the warmonger choice, but "republic" can do the same : in reality, there is as many exemple of peacefull or warmongers country with on or another. It don't matter : the gameplay is the goal with good synergies with some civics.
3- rewrite all, with a scientific approch -> design civics not following real ones, but with a position between opposites axis, for exemple, a "liberty" row where you can go from "zero liberty, all right to one ruler" to "100% liberty, no rights to no ones". You can call them "despotism" and "anarchism" but it is not necessarily. But with this way of seeing, it could be difficult to balance this with an interesting gameplay. Difficult also to choose wich axis will be used : liberty, religion, power application (from power of the mighty to power to the crowd debate)...? don't even know if all of them could have an effect in the game, because the game ins't a realistic simulation, so...
I think in this mod you choose the second path. But in my mind, for succes to this, civics should be in accordance with the ways for victory : each type of victory should make sense with some civics (and leader traits) synergy. And it's the reason why, in my opinion, the "environmentalism" civic is so useless.
For religion, what bother me in the vanilla's is the game always end with judaism or hinduism everywhere. It's absurd but logical with the game mecanism. Limit the religion to only one for each civ or so don't really solve the problem for me, it become just a sort of secondary civ flag. If religion aren't all founded at the same date, and are still link with techs, then it could be more interesting to have religions for differents victory paths : religions for warmongers, maybe christian and islam, could be use the same mecanism of mahdi in dunewars and lead to deadly crusades. Religions for peacelover open to other civs, maybe buddism and hindusim (easy to spread, don't make malus with others beliefs, for exemple). And religion for isolasionists, maybe judaism and taoism/shinto (those could erase all other religion of the cities, like quizarate or bene tleilax). Something like that. And all religions should be founded with tech in accordance to their flavor : wars ones with warmongers tech path. Like this, even military civ can succes to found a religion, and cultural civ can't, like in vanilla's, found them all in chain.
