It isn't just # of users, though - SlackOps is probably unique to Slack in that list (minus Teams, I guess) - so # of messages per month is a better metric. Not that I'm letting Slack off the hook, it still may be that their codebase and/or dev process is just nasty.