Parser performance improvement: Look into parallelising the parsing of multiple files

[Rheeseyb]

Depends on #5654

If we no longer need to keep track of the existing UIDs when parsing a file, we should be able to parse each file in parallel, so we should investigate that and see how possible that is and how much work it would take. I've suggested a timebox of 2 days here, but we can tweak that as necessary depending how performance is looking after #5652 #5653 and #5654 have been completed.