Yeah, there is something of a difficult line that varies from person to person. For my own personal take, I generally consider 'spoilers' to be something really major, such as the death of a character or a similarly major event. I'm not generally too bothered about lesser events, and I don't really bother to spoiler tag things unless I think they're really major.
The most effective criticism is certainly going to involve discussing the text as a whole, and I think the idea of spoiler free criticism is kind of silly. The distinction I think falls more on if something is being designed as a review to let people know if they might be interested in something (in which case, light to no spoilers is probably more appropriate, but with the scope of a review in general being a more limited 'this is good/bad' thing) or critical analysis of a work, which will require getting into all the nitty gritty details and involve pulling out quotes, looking at things from multiple angles, etc. The latter is going to be a far more interesting experience for the reader or listener.
At least in the case of the Red Wedding, I think it works just as well even when you know what happens (on my re-reads of the book, I actually find it far more effective than I did the first time, because of the buildup to it and all of the wrong things that you can see even if the characters don't). I don't think it's something that relies on being a surprise to be effective storytelling. It is, however, something that is intended in part to be a huge surprise, which is why I think it's the type of thing a lot of people wouldn't want to know about in advance.