The problem is that this makes having line breaks that are not paragraph breaks in the output much more awkward and I think those are much more important than line breaks that are only there in the source.

This is especially true for Markdown which is supposed to be a pretty rendering of conventions that were already common in text only communication so it's weird when explicitly entered line breaks are ignored in the output.

The significant majority of markup languages essentially treat a single line breaks as a space. HTML, Markdown, et cetera. In lightweight markup languages, you normally need a blank line (i.e. two line breaks) to signify a paragraph break.

GitHub issues and discussions are an outlier in treating them as hard single line breaks (which are not paragraph breaks).

Most plain-text communication used to use line wrapping, often not supporting lines above, say, 100 characters.

Just like typeset prose uses wrapping, because your paper isn’t infinitely wide.

Good thing about Markdown is that the lack of a proper spec means you can pick one you like (when possible). Pandoc for instance treats input Markdown line-breaks in a sane way, allowing semantic breaks to not affect the output.