I'm the author of DomTerm and the above-mentioned xterm.js PR. Both use the full UnicodeGrapheme Cluster Boundaries algorithm (https://unicode.org/reports/tr29/#Grapheme_Cluster_Boundarie...). However, I haven't seen any specifications for how wide the resulting clusters should be in a mono-space context. So unless we enhance terminals to handle variable-width fonts (which I've been thinking about), we need to took at other terminals and make judgement calls. Generally, the width of a cluster is the width of the widest codepoint. Also, I decided that a "normal" character followed by emoji-presentation makes it 2 columns wide.
In your linked-to article, you suggest 2-em dash and 3-em dash should be 3 and 4 columns respectively. That might be reasonable, but it is explicitly contrary to the EastAsianWidths specification. You also suggest that Pictographics fullowed by text-presentation should have width 1. That seems reasonable, though I don't implement that.
To my knowledge, there is no specification for monospace width. Applying the East Asian Width attribute as a proxy does not work reliably. So in my opinion, it is all a matter of agreeing on what's reasonable while keeping it simple. I.e. if allocating text-presentation Pictographics to a width of 1 is reasonable, that's what renderers should do. It's what users would expect.
In your linked-to article, you suggest 2-em dash and 3-em dash should be 3 and 4 columns respectively. That might be reasonable, but it is explicitly contrary to the EastAsianWidths specification. You also suggest that Pictographics fullowed by text-presentation should have width 1. That seems reasonable, though I don't implement that.