LU-14377 tests: make parallel-scale/rr_alloc less strict

LU-14377 tests: make parallel-scale/rr_alloc less strict

test_rr_alloc() sometimes fails with a difference of 3-4 objects
per OST, after creating 1500+ objects on each OST.  This should
not be considered fatal.  Make the test more lenient, and allow
a difference of up to 0.3% of objects between the OSTs.

Fix some code style issues in the test.

Lustre-change: https://review.whamcloud.com/48914
Lustre-commit: b104c0a27713899a4d047f56fed57c30c39b8195

Test-Parameters: trivial testlist=parallel-scale env=ONLY=rr_alloc
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: Ib6ba8c5d8e9d3245833448a52f8ed25308698a33
Reviewed-by: Lai Siyao <lai.siyao@whamcloud.com>
Reviewed-by: Elena Gryaznova <elena.gryaznova@hpe.com>
Reviewed-on: https://review.whamcloud.com/c/ex/lustre-release/+/49607
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>