I had a skim of both links (sorry but just dumping a list isnt exactly proof), and taking the first example [0], I really dont see what a blockchain brings you over a database. When you have a private blockchain, or a managed blockchain by an entity, you inherently have centralisation. What happens in the automotive case if the blockchain is forked? Which chain do I trust?

Every single use case (I have yet to see a single one, and I am) has been a use case for an append only ledger, not a decentralised ledger, and in every one of those use cases the problems that a blockchain actually solved and the advantages of a blockchain are immaterial. Can you provide one example in the two lists (which you've claimed are good examples) where the properties of a blockchain are wanted, not the properties of an append only ledger? Because if you want an append only ledger, a SQL database (or even a nosql database) is a simpler solution that will perform orders of magnitude better than 500 transactions per second.

[0] https://www.ibm.com/case-studies/renault/

None of the "research" and use cases linked need blockchains or are made better by blockchains. These are quite transparent cash grabs while hype is strong.