Seeing that IPv4 and IPv6 aren't interoperable and seeing as 75% of the internet is still running on IPv4 and assuming that the amount of v6 only hosts is absolutely minuscule, I would say that adapting and then running a bot on v6 is absolutely not worth it yet at this time.

Any reason to suspect that it wouldn't be a similar percentage as IPv4 traffic?

Only reason I can think of is it's way easier to get loads of ips to bypass rate limits with ipv6 than it is with ipv4

Which is only useful in the narrow case in which your target supports IPv6 and doesn't treat each /64 block as one address for the purposes of rate limiting. This combination seems quite rare.

Why would you think it's rare? I've seen places running multiple projects were developers were barely aware that the sysadmins gave everyone an IPv6 address, and therefore the rate limiting has to be IPv6-aware.

It's trivial to fix once it becomes a problem (at least in our case it was), but I wouldn't expect it to be an uncommon mistake.

It’s pretty easy to acquire blocks larger than /64.

if you rate-limit, you should probably be rate-limiting on the /64 prefix and those are about as hard to get as individual v4 addresses.

It would be trivial to rate-limit the whole subnet then.

Maybe, but that does not mean a lot of ISPs haven't done that.