I was e-mailed a feature request to support Tibetan for category collation, forwarding it here. The requester said they will register and CC themselves here.
Tibetan (bo) collation doesn't seem to be supported by ICU right now, but the requester found a page with instructions on how to add support for it in MySQL . I browsed around that site a bit and also found  which is larger overview and has a link to Mimer collation charts . Could we reuse something given there for ICU somehow?