- If accuracy is as accurate as your emphasizing, a ship moving wouldn't significantly reduce its accuracy (as long as the computer can calculate).
Never said it would. Processing time is the question, and frankly within twenty years we'll surpass the human brain. By then predictive firing and computer assisted/corrected aim will be child's play.
- Ships are either moving towards a certain strategic goal in the area or will move strategically around the area example: group A is fighting group B, group C who are allied with group B just appear in sensor range. If group A thinks they can annihilate group B alone and then group C alone (not both together), they would start heading in the opposite direction of C (most likely not as fast as C is approaching), so it takes a bit longer for C to arrive than if they were not moving at all.
remember this can is on a massive scale where fleet C is out of range of any attacks and even fleet A and B are extremely far from each other.
If the designers are going to be wasting processor cycles with unit AI like this I say forget it.

The limiting factor on realism is computing power, and to waste it on something little more than a gimmick (like this thread) is not a good sign.
- Would take more energy for the ship to fully stop each time it starts shooting, than just continue moving and slightly changing angle/velocity.
*cough* Antimatter, jump-drives, and weapons that consume more energy than retrorockets argue this point to nil. If you had a choice to let yourself drift into fire you would not do it.
anyway most of these games are extremely unrealistic

.
To a point they have to be. The game is already overflowing emulated 64bit numbers if I read one of the preview right. Only so much can be done with today's computers, and there's no guarantee that realism is fun. Physics is right up with graphics as the heavyweight load contender in a game.