'right tools for the right job' seems like just re-stating the original point in different terms, how right a tool is depends on the actual context a big part of that context is culture, preferences, politics etc. not just engineering concerns
If “right tool for the job” is too hand-wavey, maybe “optimal tool given the problem constraints” is a better phrase.
E.g. I’ve seen proposals for “credentials on the blockchain”, but if you assume that a credential issuer has a known public key (as these proposals do), the issuer can just sign a credential, no public ledger is needed. The complexity of using a blockchain for this is just waste.
sure, but the problem contraints arent reducible to engineering, so in the example you're giving an established centralized authority can profit from selling credentials using only a key signature, but a new decentralized bottom-up credential system needs a blockchain and smart-contracts to do it ...
The problem is that most such schemes end up requiring both a blockchain and an established centralized authority, because the blockchain doesn't actually solve the problem its meant to solve (since the only problem it solves is distributed consensus on the order of events in the blockchain - no more, no less).
I think it does, because of its complexity, it's more of a set of technologies than just a software solution. Preventing double-spend is just the main use, however in order to do this a whole game-theoretical economic/technological system is required and that also solves and/or creates other problems. For cryptocurrencies, DAOs and similar systems economics is a security concern (wrong incentives will make it possible for an someone to game and/or break the system).
The point is that blockchains don't provide you with anything other than that "main use". For any other kind of feature you need to add something else - and if you're adding a centralized trusted authority (which is by far the easiest way to do anything), then you can get rid of the blockchain altogether.
