I agree with a lot of the people who compare it to the idea of chess with simultaneous moves. There is a massive amount of waiting to the last moment of a turn to move a unit, then moving that unit immediately at the start of the next turn.
If the game wants to insist on simultaneous turns, one good option is something akin to MOO with a conflict resolution phase.
Basically how this works is that the orders happen based on actual turn order, but if something changes (for instance, an attack becomes a move) then in the conflict resolution phase this units order can be changed to whatever action you choose based on the new game state. Since actions are executed in normal turn order, you wouldn't have any more information that you would have in the normal (turn-based) game state.
I agree the bigger difference is that humans are better than AI, but when I played competitive online back in the day I used to insist on non-simultaneous turns (I took up non-video game competitions so haven't kept up with Civ as much). There is just way too much double-move exploitation, and while it isn't quite as problematic as double moves in chess, it isn't too far off either. While it does solve the problem of games taking a long time, it does so at the expense of changing the nature of the game.
One practical solution might be the following: Y second build phase (can not move units), X time unit move phase per player, non-active players can change build orders). This gives you a time of Y+2X in a two-player game which is the most typical case for competitive multiplayer, and X is typically a lot smaller than the normal turn as gamey exploits results in turns taking close to the maximum timer length.
Another practical solution is the EU style "Real-Time" where the economics of the game are basically turn based with turns happening every clock cycle (in EU one month), and the military interface (unit movement ETC) takes time based on terrain cost and unit speed so that an X speed unit will arrive in a square that costs say Y days/unit of speed to move into will arrive Y/X days later. This does lead to a few exploits with regards to things like unit attrition (not implemented in civ), but as a practical matter in no-pause games timing exploits is difficult.
In Civ for example you might have time passing at say 1/40th of a turn per second if you wanted 40 second turns. This would mean that the game clock would move 1 year per 2nd.. and a unit executing one turn of movement will finish moving 40 years after they started moving.. so If I start moving in 3660 BC I arrive at 3620BC.. and can then start my next movement which would complete at 3580BC even though the normal turns at this point are 3680 3640 3600 3580. The economic updates still happen at the normal end of turn years, but the unit exploits are basically gone.
I much prefer Civ as a game to EU I think Civ is a lot richer, but I strongly feel the multiplayer in EU is much more interesting. I'm not opposed to rush strategies, in fact I dominated at them for Civ3, but move exploitation is so important that in competitive multiplayer it is an absolutely necessary skill, and there are a sizable number of people who will avoid competitive multiplayer because how dumb the adhoc solution makes the game.