[qw]

It's similar to how some development choices are made too. Examples: Rewriting the system in a new language, change a framework or create your own from scratch.

I have fallen into that trap myself. I have read about some exciting new tech in a blog and want to try it out in a POC. There may be some excitement around it and suddenly it becomes a MVP and placed into production. Eventually it evolves into "legacy code" that developers hate to touch. The right approach is sometimes to step back, and rewrite the POC in a language/framework that can be supported by others.