Running SQL Profiler against a production server should be a fireable offense. It puts horrible load onto a server, especially if the person hasn't been careful about filtering their trace. Whether or not it caused the high batch requests or not, it probably caused the performance problem.

Suggestion, revoke that developer's ALTER TRACE rights (and make sure they don't have a sysadmin login) and educate them about server-side traces.

I've crashed a production server by starting Profiler with a light-weight trace. Crashed as in the server froze and the cluster failed over.