The thing is even if SpaceX gets actually cost down to ridiculously low prices it's still going to be extremely expensive and very very hard. We have practically speaking no experience manufacturing in space, the only thing that's been made have been small test quantities of ZBLAN fiber and until we get asteroid mining and refining setup the extra cost of lifting all the raw materials is going to make manufacturing in space only viable for a few small things with high margins of return.
You'd still need it LEO at 500km varies between 10-700 nPa and UHV is a max of 100 nPa. It's too variable an environment to really develop chips in due to space weather and the variability in the extreme thin reaches of the atmosphere in Earth orbit.
The only think I really know of is ZBLAN fiber because that has a definite benefit to space manufacture and a large profit margin to justify the lift costs of the base material. I think until we get space refining figured out it's going to be hard to manufacture anything because of the high lift costs. Modern manufacturing also depends on a large amount of equipment and chemicals that will have to be redesigned for zero-G [0] and whole supply chains of chemicals either lifted dry and recreated or created from scratch in space.
I think the first things we'll really make in space will be super structures and pressure vessels because those are the simplest things to make and the materials don't require a massive supply chain. We could even gain a lot just by lifting up generic stock from Earth (metal powders, sheet stock that could be formed into tubes, or just a lot of generic tube stock [1]) and learning how to assemble them via machine in space. It would let us build larger stations or telescopes than we could launch from the ground.
For example if we build a super structure in space we could launch a lot of large inflatable modules to fill it in with that would be mechanically simpler and lighter because they don't have to bear and transmit the thrust force just support their own mass under thrust.
[0] Which increases the amount of benefit needed to justify redesigning an entire chip fab for example to function in zero-G.
[1] Though lifting tubes has the problem that there's a lot of wasted space in a stack of tubes which is one of the problems we're trying to avoid by building vehicles in space to begin with.
imo the most lucrative thing to manufacture would be parts for larger orbital stations and for satellites/rockets. The most expensive part of space is getting there - if you could manufacture the payload (aka body of the rocket) in orbit, there would be no need for a stage 1 or 2.
