Introduce parallelisation for finding projections, extracting metadata and saving the stack

[ptim0626]

Currently the time taken to find projections, extract metadata and save the stack does not scale well. It is desirable to have some parallelisation for those steps.