Since there are competing concerns, we have to decide how to weight them. For this, it is helpful to know what one is optimizing for. On HN, we're optimizing for intellectual curiosity [1]. Common/generic annoyances are generally less interesting than specific new information in an article, so we favor the latter over the former.
It's not that the other concern is invalid, just a lesser priority relative to what we're optimizing for. Optimization is knowing how to rank valid things. It's easy to drop invalid ones.
The current thread is an example. The highest comments after I marked this subthread off topic are https://news.ycombinator.com/item?id=30386892 and https://news.ycombinator.com/item?id=30384256. It's better for either of those to be the top subthread instead of https://news.ycombinator.com/item?id=30384311, even though that popup is annoying.
You can't go by upvotes alone. Some categories of comment routinely get more upvotes than they should, relative to https://news.ycombinator.com/newsguidelines.html. This is a weakness of the voting system [2].
[1] https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...
[2] https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...