I agree. I am not suggesting that the SQLite team doesn't know how to make the technology better. Just that they aren't/haven't. Backwards compatibility is a good reason not to.
My original comment was contrasting craftsmanship and utility, since both are somewhat prized on HN, but they aren't the same thing at all. Look at a system like Wireguard. A huge amount of small decisions went into making that as simple and secure as it is. When most developers are confronted with similar decisions, they perform almost randomly and accumulate complexity over the long tail of decisions (it doesn't matter just pick a way). With Wireguard, every design decision reliably drove toward simplicity (it does matter, choose carefully).
I'd say they're prized everywhere, though "craftsmanship" is really subjective. and the HN I usually [edit/add: see] seems to have more a meta of "criticize anything someone tries to build, and rave about IQ" tbh ;)
SQLite works and I don't have to think about it why it works (too much). That is IMO a true hallmark of solid engineering.
Long term compatibility (i.e. prioritizing the needs of users vs chasing inevitably changing ideas about what feels polished or pristine), near fanatical dedication to testing and quality, and sustained improvement over decades - these are the actual signs of true craftsmanship in an engineering project.
(plus, I don't agree with you that the storage layer, column format, or SQL implementation are bad).