here's a bit of confusion some people might have though:

Emoji (represented as such) are not a part of the Unicode Standard, really. Unicode 10.0 basically says "emoji exist and it's okay to display some unspecified set of characters as emoji" and that's it. Emoji are specified through a Technical Standard which is basically a BCP stating how most people in the industry currently handle emoji display, and what you should do to ensure interoperability/reduce confusion.

the emoji standard basically consists of:
- here are the accepted techniques for going from the set of characters provided by the Unicode Standard to emoji representations (variation selectors, ZWJ sequences)
- here is a common set of characters that everybody should do this for (but you can do this for others if you want)

the goal is to ensure apple and facebook and google and others all agree to do things the same way, so that users don't suffer unexpected effects.

unicode used to have a thing (idk if they still do) titled "the future of emoji", which talked about the very-long-term plans for emoji and images-in-text. it basically said:

1. For CLOSED PLATFORMS, and RICH-TEXT applications, shortcodes and image embedding was a better solution
2. However, for PLAIN-TEXT applications designed for GENERAL INTERCHANGE (cf. email, documents, archives) there would ALWAYS be a need for Unicode emoji to ensure interoperability

and the *entire* unicode approach to emoji is based on this premise: how can we encode things in a plain-text format that will outlive any specific implementation or indeed even the Emoji Standard itself (remember emoji ≠ unicode)

when it comes to emoji in Mastodon, really the question comes down to whether you see the platform as (a) closed or (b) open, and (1) rich-text or (2) plain-text. For a closed platform, proprietary and unstandardized shortcodes are the best solution, but these don't work for an open one. For a rich-text platform, embedding images is fine, but that's impossible for a plain-text one.