QES while I like your idea of making religion adoption based on alignement and not vice-versa, I don't think that tying alignement on civics would work, either. This because civics are not balanced toward each other, so the choice of them is not simple flavor/preference but a matter of strategy. This means that if you like a certain set of civics you will always (or almost always) play a leader of a certain alignement, indipendently fom the civ etc. The AI would probably follow an even more predictable path than now. The weakness of AI is exactly its predictability.
While I agree with the AI and predictability i am a bit puzzled by your other assertion.
Specifically - when people understand that civics limit or influence what religions one can take in a game - would then that still not be a part of strategy?
If religions had no differences between one another - like in vanilla civ - then you'd be quite right. However, since each religion has its pros and cons, and players will oft "seek" a religion from the start (or at least consider some more strongly than others) having civics impact alignment, and therefore the ability to choose certain religions would GIVE the civics each more importance and aspects IN strategy.
If I pick a leader who is realatively evil, (say -3), and I pick a few civics that are "good" and my total comes out to 0 - but I want to get the "ashen vale" I am now going to have to make some hard choices. Given that these choices are equal across the board for everyone (alignement to religion) they become part of the strategy. Also - if you want to play the Elohim and ashen vale, there'd be some necessary choices to make to becoming "evil" enough to be heard by the demons.
There is another side to this that we could consider.
When civics lead to alignement, and alignment leads to religion - then each step along the way can be more independantly adjusted.
What I mean is, that when alignment becomes a buffer between the two catagories - then the realative balance changes. Kael has often remarked that he likes the wholistic approach to balance instead of individual schemes.
When civics and alignment and religion are tied at least loosely together, then the flexibility with any given mechanic is given wider berth.
I am overcomplicating my case - my appologies.
Simiply put - I believe that having these things be interconneted would in fact
increase strategy and importance of each civic.
For example - if a civic is sucky (or at least you never ever like to play it) but it would allow for you to pick a religion you like because of a alignment modification - suddenly its value is increased.
If, on the contrary, you love a civic - but choseing it would suddenly negate your ability to enforce a certain state religion - its relative value would decrease.
This alone makes for easy methods of increasing "under used" civics, and mitigating "overly used" ones. Granted, this might require an overhaul of the entire civics system - but from some of the things I hear, that may be a necessity anyway.
-Qes
Edit: In my head i see each catagory of civics as having a diverse representation of alignements.
Perhaps if there are 6 options in a catagory (all technolgoies given) then perhaps one of them may be an extreme choice (+-2) 3 of them a minor choice (+-1) and the last 2 have 0 impact.
This would do two things - A) civic choices and not leaders would have the ultimate impact on final alignement. B) As technologies progress, alignments may shift and flow throughout the game (creating an automatic ebb and flow of good and evil as time passes - which sounds very cool to me) finally - the technologies and therefore specific civics for each religion would strengthen anyone's alignement who sought that religion beyond normal - thus allowing more choice among the other civics.
I dont know how FoL's tech would "make one more neutral" but I'm sure there's a way. Perhaps it would make one's leader mod value 0 (as long as it was selected).
-Qes