the -w option in traceroute seems to be broken. when tracerouting to a non-responsive host the traceroute doesn't stop after the time wait period, it goes on for ever unless a user kills the process. this seems to be the same bug that is in the standard rh6.1 ping. Ping was fixed in rawhide 1.0, traceroute hasn't.
Addition to the above statement: the -w option in traceroute is not equal to the -w option in ping. it's equal to ping's -i option. which means that traceroute is lacking the -w option that ping has, though it would be a VERY good idea to have ping's -w option in tracerout for automative purposes. your choice whether this remains an open "bug" report.
You're correct, ping and traceroute and ... should have similar options because they have similar purposes (e.g. debugging networks). I'm gonna mark this is deferred, we'll get to this someday ...