I don't know what their reasoning is, but I suspect it's a mix of everything discussed on this thread: ancap/right-libertarian ideology, an affinity for sexy tech regardless of usefulness, faulty reasoning by historical analogy, reality distortion due to having made money off the crypto bubble, and a bit of sunk cost fallacy.
I am willing to be proven wrong. Show me growing uses of cryptocurrency that are not one of: niche wire transfer applications, black/grey market commerce, cryptocurrency aficionados nerding out, gambling, speculation, fraud, or solutions in search of a problem. Those are all (except the last two) real applications to varying degrees, but they're not large, rapidly growing, mainstream, or of that much interest to business customers.
In other words show me something someone that doesn't care about cryptocurrency would use. The Internet in the mid-90s offered a ton of things people wanted to use even if those people didn't know what an IP packet was.
I have asked the above in various forums, Twitter, etc. for years and so far nobody can produce anything. The closest I've ever seen is this:
It's the only ICO project I know of that has actually shipped something usable and possibly useful. Still that's one singular example and it's not like they're going to displace Amazon S3 or Backblaze any time soon.
Is it a good thesis? What's your criteria? By any scientific criteria, of course, the thesis is asinine. However, if the criteria is other VC theses, then it is an adequate thesis.
And numerology with suspect numbers at that; 1978 was the year after the introduction of the first viable personal computers, 1993 was, er, Mosaic came out I suppose, but it was not the year the web showed up (that was earlier) or the year that it started to see widespread adoption (that'd be, er, 95 or 96). 2007 was the year that smartphones became mainstream, if you don't count blackberries or other consumer quasi-smartphones. This all feels very cherrypicked to fit the thesis.