When you enable RUM capture limits, there is no way to ensure a specific user is tracked. The way I've gotten around this in the past is turning on 100% capture during the hours of the test and then disabling it. This takes a much more active approach to monitoring around the test, but ensures the whole test is captured, as well as the specific user. In my case we only tested for 3 hours, hopefully you don't need to run this test for days.
The other option could be to enable capture, force that test user to execute (if possible) and then after execution is complete, reduce back to 30%.
Hope this helps!