So, so far, this is accurate. If Musk can mass-produce TBMs and, *more importantly*, break the local construction monopolies, he could cut costs significantly.
However, so far he isn't doing anything which would break the local construction monopolies. The really really overpriced and delay-prone part is NOT the TBM part of the construction; it's the surface access points. He needs something which will replace THAT in order to break the local construction monopolies.
If he has a fast TBM and the people digging the access point are from the local construction mafia, he's going to make essentially no progress on dealing with the cost issue. This should be obvious, right? I would really like to get Musk to lunch for an hour to point him in the right direction. Automated bulldozers and backhoes would probably do the trick.
The "You never know what you're digging into" remains a problem for the tunnelling portion, and there's no way to get around that. A combo TBM which could handle all three of the common types of ground would be great, but I'm not sure Musk even realizes that there are three different types of TBMs yet.
(P.S. You can't make the TBMs any more reusable than they already are; they are very reusable. The problem is fundamentally that the cutting heads wear out as they grind away at rock, and they *must* be replaced: they are a wear item, like tires, or rocket fuel. )
(P.P.S. If you're going to mass-produce a standardized TBM, making it slightly larger than London Underground tube size -- as Musk has proposed -- is a totally reasonable choice of size, as it works for most subways, but calling this larger diameter "smaller diameter" is simply false. I believe Musk made false statements about this due to gross ignorance... and that's not a good sign.)