Well that is interesting...
It looks like a small number of assumptions have conspired against us.
A project lead can initiate a committer election. The election is considered complete if all committers have voted. The logic for "all committers have voted" comes down to a simple count.
By way of rambly background... we've always treated the project lead and committer roles as distinct. That is, the project lead role is not some sort of "super committer", but rather a distinct role that comes with distinct privileges and responsibilities. In practice, it is almost always the case that when somebody has the project lead role, they also have the committer role. While I'm at it... the way that team roles are set up on GitHub and GitLab, a project lead actually is a "super committer" with regard to privileges on repositories. This has led me to the conclusion that any assertion that the project lead and committer roles are distinct is actually meaningless and, well... pointless. A project lead is for all practical purposes a super committer and the current means of reflecting that in our database is to have every individual who holds the project lead role to also be a committer.
All of this is to say that I'd actually thought that we'd resolved the extremely rare cases where we'd had a project lead who was not a committer. Obviously we missed this one. Making sure that the project leads are consistently a proper subset of the committers works around this problem.
Still... I'll open an issue.
Thanks for stepping up, Ed. This project needs some love.
Wayne