. . . you start new huge campaign, steamroll over hapless aliens for a while and then get bored and abandon the campaign just to repeat same sequence but in different settings. . .
I think most of those (at least for the C# version) can actually be attributed to Steve making major changes to the database which require him to flush everything; rather than boredom.
Good and true point.
However, Steve himself differentiates between test and regular campaigns. Please notice, these 4 campaigns are NOT classified as test ones, which kinda renders that point moot.
All i wanted to point out is:
1. better to have less elaborate campaign setup than opposite, so as not to eat up too much time on early solo game (arguable, but see e. g. W40k)
2. AI is apparently way too weak in his campaigns (95% steamrolling, human brain is still OP AF)
3. expanded point 2 - as with most strategic games, one has to impose (many) house rules on himself to give AI chance (otherwise its back to steamroll)
Like i said, its my opinion/advice to make his campaigns AARs more interesting and make less of aborted ones. Much of this stems from my own ongoing search for Aurora campaign balance. Still trying to find a way to have plausible setup with competent opposing AI without getting 15 minutes long month turns or infinite 6 hrs increment hell. . :-\
YMMV, of course
unrelated question, is there any plan to re-balance pops for bit more realistic values? And ideally have AI prioritize to re-settling overflowing pop from high difficulty starts?