I think by adding maintenance costs to starbases it would reduce, rather than increase, starbase/shipyard spam. Starbases would be most useful to provide navigation routes and "internal" starbases would follow long after that, perhaps only being justifiable in one's core worlds where the research/manufacturing/wealth dividends would outweigh the maintenance costs.
I'm still concerned about how the AI would cope with it tbh; their SB placement tends to be in the core or on resources. IT doesn't think in terms of 'I'm out of range, I should put a SB right there." More 'I've built a constructor! Yay! Let's roll a dice to find a planet or a resource to plonk it next to. I wonder what type it will be. Let's roll the magic dice and see!".
- Give one's home world 100 base population (you did say you scaled that up by 10)
Breaks custom races. Otherwise I'd already have done this. For some unfathomable reason, SD decided to set starting POP on a per-race basis. The same is also true for money. I've honestly no idea why SD did this. It genuinely makes no sense - you can mod it in the xml, but you can't mod it in the race customize (which then bakes it into an un-editable Faction.xml file). Any benefit from it could easily be captured using a 'base' setting in globaldefs, and then modified in the race defs.
- Return to 1 total production point per citizen
Not sure why you'd do that, tbh. It would speed the game up - but probably way too much if you have 100 production on day 1. And anything else would mean a big upward price re-balance across the board, which would simply cancel out the effect.
- BUT you should attach a maintenance cost (not just an opportunity cost) to military production. In real life it is rare that anyone would get to 50% military production; the only time that has been reached was during World War II. Nowadays, even the US's 5% of GDP going to the military is considered extreme. Military production ought not to go above 10% ordinarily without bankrupting your treasury.
Not sure this can be done, aside maybe from very high shipyard maintenance levels - though again, the AI doesn't pay a blind bit of notice to anything I tell it with regard to shipyards. Also, let's not forget that 'military' production basically means 'everything the government does that isn't buildings'. I prefer to think there's a sizeable private sector going on behind the scenes, and you're merely playing with the taxed quantity.
Even with the current checks you have placed on colonization, some AIs are expanding too rapidly. Within the first 150 turns I was only able to grab half of one large cluster, while the Iconians (by geographic chance) had taken 3 large clusters and most of the relics around them by that time. Expansion ought to be something that is carefully considered even within a cluster and difficult without.
I think what you're aiming for would tip the balance too far in favour of Tall play over wide. While there's benefits to this - namely, that starting position is considerably de-emphasized versus good play - there's also problems with it - it will take forever for resource scarcity to force tension into the game, and the AI really isn't built for it. I'll bet that you were still reasonably close in the power ranks to the Iconians, despite their enormously large number of worlds.
I generally avoid cluster maps, precisely because random chance can screw you on turn 1. I had a bad experience with GC1, where my first game on a cluster map was me in one corner with 3 stars, and by the time I could reach anyone they were all massive and all the planets had been taken.