perf(core): persist workspaces in parallel

[merceyz]

What's the problem this PR addresses?

Yarn persists workspaces one at a time which with 1000s of workspaces noticeably slows down an install.

How did you fix it?

Persist the workspaces in parallel.

Results

Tested on the DefinitelyTyped repo with its 8629 workspaces (matching only types/*):

$ YARN_IGNORE_PATH=1 hyperfine -w 1 "node ./after-plimit.cjs --mode skip-build" "node ./after.cjs --mode skip-build" "node ./before.cjs --mode skip-build"
Benchmark 1: node ./after-plimit.cjs --mode skip-build
  Time (mean ± σ):     16.106 s ±  0.605 s    [User: 13.252 s, System: 9.803 s]
  Range (min … max):   15.090 s … 17.147 s    10 runs

Benchmark 2: node ./after.cjs --mode skip-build
  Time (mean ± σ):     16.644 s ±  0.635 s    [User: 13.874 s, System: 10.015 s]
  Range (min … max):   15.932 s … 18.146 s    10 runs

Benchmark 3: node ./before.cjs --mode skip-build
  Time (mean ± σ):     17.390 s ±  0.641 s    [User: 13.684 s, System: 10.219 s]
  Range (min … max):   16.582 s … 19.000 s    10 runs

Summary
  'node ./after-plimit.cjs --mode skip-build' ran
    1.03 ± 0.06 times faster than 'node ./after.cjs --mode skip-build'
    1.08 ± 0.06 times faster than 'node ./before.cjs --mode skip-build'

Checklist

• I have read the Contributing Guide.
• I have set the packages that need to be released for my changes to be effective.
• I will check that all automated PR checks pass before the PR gets reviewed.