Try to clarify things in the comment too.

Also add rounding support to tv_mdiff().

A) We were considering a circuit had timed out in the special cases
where we close rendezvous circuits because the final rendezvous
circuit couldn't be built in time.
B) We were looking at the wrong timestamp_created when considering
a timeout.

Based on irc discussion with arma.

Mostly by storing the timeout as milliseconds and not seconds
internally.

until mike figures out some other way to make them pass reliably

Don't discard all circuits every MaxCircuitDirtiness, because the
user might legitimately have set that to a very lower number.

Also don't use up all of our idle circuits with testing circuits,
since that defeats the point of preemptive circuits.

[Do not take conflicting change to rendclient.c]

And also the number of recent circuits used to decide
when the network changes.

We want it to be under our control so it doesn't mess
up initialization. This is likely the cause for
the bug the previous assert-adding commit (09a75ad3) was
trying to address.

Got a negative timeout value on startup. Need to narrow it down.

Also use bin midpoints for time values.

It seems to fire because of precision issues. Added
more debug info to the warn to try to figure out for sure.