I mean, his main argument is that costs that revolve around the TBM do not make up enough of the total cost of tunneling that TBM and tunnel diameter based cost savings can generate the costs savings that Musk wants to achieve. His primary data with regards to TBM and diameter costs comes from Europe (which ya, is not USA, but closer than Asian countries), and comes up with a rough upper bound of 50% of total tunneling costs in the LHC were associated with the TBM (this is bound to be lower for any transit application, that isn't a point to point tunnel due to additional group level access and extra boring).

If TBM costs are 50% of total cost, no amount of TBM cost reduction can achieve the 90% cost reduction goal - it clearly has to be combined with other strategies.

I believe it's clear that the author believes that these other strategies (all of which is probably project management based strategies) should be engaged with first, before falling down a path of throwing capital and R&D at cost reduction mechanisms (reduction of tunnel diameter for example) that may have significant technical and long-term short comings.

I mean, the entire blog post is just telling Americans what Americans kind of know - Large American government contracts are often poorly managed and executed (for whatever reason), and that reaching for a technological magic bullet is unlikely to achieve what we want.