I agree with you. The problem is that it's impossible for a chess engine to know what's easy to see or hard to see. All it knows is that the eval was at +2 and now it's at -7, so it shouts blunder. It doesn't matter if it would take a GM 10 minutes to figure out why, the engine will still call it a blunder.
When I was new chess.com didn't exist, so I didn't have to worry about it 
But still it would be frustrating to see the annotator explain "this alternate move doesn't work because blah blah blah" but he didn't explain why the super obvious win of a pawn wasn't good. So as a beginner it was frustrating because it's just a mystery why the GM didn't win a "free" pawn.
One thing you can do is try to point out a few of the biggest mistakes, and a few of the best moves yourself, with no engine help. Then post it in the chess.com analysis forum and get human feedback. That sort of feedback will make more sense, and you can ask followup questions when people are unclear.
Can anyone define "Inaccuracy", "Mistake", and "Blunder"?
In analyzing games, I sometimes have many "Blunders", and I really don't get the definition. It seems like a "Blunder" should be fairly clear - like a low level player would look at the position and say "Oh, that's what I missed" or right after the move they say "that was dumb" (which happens to me ALL the time)...
But it seems like the analysis is more along the lines of "After..., then..., then... and you arrive at a position where you still don't see the problem." How is a low level player supposed to see all that? It definitely seems like "Blunder" is not a relative term, and just makes learning frustrating.
I get that I make lots of "Mistakes", and TONS of "Inaccuracies" but shouldn't a "Blunder" all but cost you the game?