To me, it just seems like the author wants to hoard their data, not make it useful. Making it useful (the first part of which is making it queriable) involves making some kind of interface to access it, and some code to parse it. In my own case, I find that just dumping JSON docs around the place gives me an unmanageable mess of unstructured data that I never come back to. If all you want to do is to hoard your data, then I guess that achieves the purpose... But if you want to do something with it, at some point you have to make sense of it.
Doesn't mean you have to go and fully normalize everything and make a production-level database!
Doesn't mean you have to go and fully normalize everything and make a production-level database!