Joe suggested that this waiting may be for a lock that wasn't released when the process crashed.
My guess is that the inspector that crashed might have a lock on something (local storage? sql storage?) during the time it takes crash reporter to create a symbolicated stack trace, and so a new opening inspector might be blocking on access.
In order to find out what is really happening, we just need to get a quick sample of the new web process while it is trying to open the new inspector.