No, not really.

Also, making a "competitive" difficulty level is an extremely high challenge in terms of programming the AI behavior, and balancing it against a very diverse playerbase.

This. The problem is, a lot of the decisions we make are easy for us. We sort of see everything at once, we can just say "oh I'll stay a little longer to destroy this capital, then I'll retreat" or whatever, but when you're trying to code it it doesn't work that way. Everything the AI does has to be calculated and quantified. In the above situation, the AI would have to query the strength of its own fleet, enemy's fleet, then check the current hp of every enemy capital, figure out the estimated time it'd take to destroy it (if you want to make it really smart, it'd also have to look at the number of enemy support ships able to repair it), do some number crunching to see how long it can stay before it loses its most damaged capital and if it's enough time to nuke the enemy's.... and it just piles on from there.

Whereas a human can just glance at any given situation and absorb all the info, the AI has to constantly keep track of it, which often means re-doing the same calculation every so often. The more of these things you have, the more bogged down the CPU becomes until the only thing it can do is run the AI calculations

Oh don't get me wrong, I wasn't saying there's no room for improvement.. just that some of these things sound easier than they are to do.

However they did say in one of the early post-release interviews that CPU was a big limiter for their AI.

The couple of times that I've seen the sacrifice issue was simple. The system was not the AI's homeworld, I had 4 capital ships (doing an achievement or teaching friends how to play) and the AI had a single cap with less than 10 frigates (skirmishers in the most recent example). The frigates did not run until the cap ship was destroyed. I have also seen the converse where a large number of frigates ran when staying could have really hurt my defenses. So I see both sides of the issue and would appreciate a better more intelligent AI.

Another example of potentially better AI play is setting defenses within their owned systems. I rarely see repair platforms and the weapons platforms seem scattered and non-supporting in terms of field of fire. There are tons of posts here that talk about overlapping repair platforms and locating defense platforms with 1-2 hanger bays to buy time for reinforcements to arrive. Yet even that kind of deterministic coding is missing.

I'm a long time Stardock player and all of the GalCiv games had some really nice AI and I didn't have to wait even seconds for each AI player to complete their entire turn to decide and make strategic decisions.

I know, however I'm also not asking for the intelligence of a grandmaster chess game. Also there is a lot that can be done in a background worker thread (not read as multicore) that can aid in the continual evaluation of combat situations as they occur.

Just so everyone knows I really like this game and see playing it for a long time to come regardless of these kind of changes. I've even purchased entrenchment even though I'm unlikely to install it, because I want to assist in supporting the company and continued evolution of this game. I've just about dropped all of my other games in favor of this one so I can get in 20+ hours/week.