Ask HN: Technology: how do you separate signal from noise?

For example, when I hear some Erlang folks talk about Erlang powering x% of network infra, the best concurrency, distributed choices, or Clojure people speaking about how they are able to focus primarily on problem complexity, and how OO is/was a big mistake, then unless I try it myself, it feels like Spring's war on enterprise java's complexity argument.

I am not saying claims like these are misplaced/ hyped. But how do you efficiently separate hype from crux. I had to listen to Rich H, Joe Armstrong talks, read Elixir book by PragDave. Defaults sometimes don't help: for example, I generally ignore all software craftsmanship debate because I have formed some opinions about it. But in the new areas what would help me separate signal from noise? Perhaps I got burned by Agile dogma and so any new claim seems to me like sound and fury signifying nothing. What can help me decide merits of things like these efficiently and separate those from hype/dogma?

Thanks.