[crystalis]

I don't think they allow, expect, or calculate an unbounded positive as t->infinity. I don't expect we'll get caught in a local minimum - we've done a decent job breaking out of them when the individuals involved haven't had nearly as much at stake...