Why these three encodings?

[domenic]

The README makes the claim they are the most common:

This is a first pass with just the 3 most common encodings. Support for "base62", "base32", and various other encodings can be added after initial consensus and implementation.

Is there any way to back this up? Apparently there are a lot: https://en.wikipedia.org/wiki/Binary-to-text_encoding

[lucacasonato]

These three encodings are the ones available in all of the environments mentioned in the proposal (Go, Dart, Node.js, Deno) out of the box. This suggests that they are the most commonly used encodings.

[domenic]

Elsewhere, @bakkot says that apparently Node.js's Buffer.from(x, "base64"), as well as CSP, support mixing base64 and base64 URL in the same string, when decoding. That might be an encoding worth supporting as well.

Cross-linking to tc39/proposal-arraybuffer-base64#7 (comment) for related analysis.