I agree about your first point: acknowledging only trust elimination and denying trust minimization is a simplification from my side. However, I have not seen successfull examples of "trust minimization" so far though it is always suggested in the context of blockchain. Maybe the reason is that the part on which the trust can be minimized by BC is usually not the weekest link.
I also agree that BC could prevent some percent of deed frauds. However, if you consider my argument about not decentralized example land registry BC maintenance you will agree that in many cases this may leed to a forgery, not possible in Bitcoin blockchain.
I completely disagree on a tokenization part. Unless we have a well-described and tested business model. these are more dreams and handwaving than a real use case. It is easy to say "tokenize this and that", the devil is in the detailes.