Fixes #28290
These fixed timeouts aren't affected by the timeout_factor option and can therefore cause timeouts in slow environments.
They are also unnecessary for the test because they measure the wrong thing:
While there is an internal waiting time of 60s within ThreadOpenConnections (beginning only when that thread is started) for fixed seeds querying, the timeouts here don't measure that but the time from startup until a debug log message is encountered, during which many other things happen in init, so they don't make much sense to me in the first place.
