Making a decision on DB is somewhat foundational for a lot of system architecture, in that you expect to be able to use it for years. It is not like some UI framework you toss out every 18 months.
So the benefits of the new hyped thing may be outweighed by the unknowns & risk of flaming out.
This is the kind of hype cycle that gives me pause when a new hot thing like DuckDB which actually ticks a TON of boxes for me, but has attracted some of the usual suspects in my work network that I consider to be contra signals.
https://gist.github.com/cpursley/c8fb81fe8a7e5df038158bdfe0f...
I mean, its a lot more expensive than Postgres for very very little benefit. Like what exactly is the value proposition? Is it saving developer time? faster? I guess none of those
These days a I am not so sure what the case would be other than legacy or very occasionally some feature only SQL Server has. AD would be a big one.
Today I have no idea about SQL server but if the support is as good today as it was back then it can be worth it.