hyperion startup uses the wrong startup log file for the supervisor, blueapi startup timeout is too short.

When the supervisor is started, the startup script uses the same startup log file for both blueapi and the supervisor, meaning that one gets overwritten.

They should use separate startup log files.

Also the startup timeout is only 30s which is sometimes too short esp when the filesystem is slow and blueapi files are not in the cache.

## Acceptance Criteria
- Timeout is extended
- Separate startup files are used.

Currently hotfixed on beamline

Locate the startup script used to launch blueapi and the supervisor, then inspect how it sets the startup log paths and the 30-second timeout. Use the acceptance criteria as the check: each process writes to a separate startup file and the timeout is extended enough for slow filesystems.

Written by the indexing model from the issue text.