[BUG]: McStas PowderN use of cif2hkl for cif-files is semi-buggy

[willend]

Describe the bug
A clear and concise description of what the bug is.

For a long period of time I have seen intermittent issues in the instrument Test_Powders that take cif/ncmat input and tests across PowderN, NCrystal_sample and Single_crystal.

(PowderN today take all of laz/lau,cif, NCrystal cfg-strings and actual NCrystal ncmat files. A similar development is planned for Single_crystal wrt. NCrystal support.)

The issues

• Occur on all platforms, but more often on Windows and macOS than Linux
• come in thee form of either
  • a full crash (detector value -1 and non-zero exit value
  • numbers that are 25%-200% off (and a non-zero exit value)
  • numbers that are 25%-200% off

After some investigation it seems the errors mostly/always come when a cif file is used, i.e when cif2hkl is used as an external process to convert cif -> hkl list.

Workarounds

• Suppress use of cif2hkl in Test_Powders #2634 is under way and will temporarily disable %Example use of cif-files.
• For most of the cif-files we ship, an ncmat equivalent exists (see nctool -b for stdlib:: list or convert e.g. a COD cif-file via e.g. nncrystal_cif2ncmat codid::1000236 # Na2Ca3Al2F14 -> autogen_Al2Ca3F14Na2_sg199_cod1000236.ncmat

Expected behavior
Should rather "just work" ;-)

Additional context
Deeper investigation of underlying cif2hkl / CrysFML Fortran code will come -> issue on relevant repo