In my experience, saying "that won't work" as a direct response to anything a peer proposes is too early (or someone even resembling a peer).

It's not that you need to say yes to things, it's that you need to sit with it a minute, and challenge your objection internally / try to creatively get around the fatal flaw you perceive, before voicing your opinion. Sometimes you'll actually be wrong, other times you'll invent something that works in response. Sometimes you'll be right, and you can raise the objection after you've put in a little internal work.

This isn't the case if, say, someone suggests something as outlandish as "we should just invent a new compression algorithm that works on all possible inputs." But then, that's not something a peer or someone resembling a peer is likely to say.