fix: don't enforce identical df index on {obs,var}m

[ilan-gold]

See scverse/anndata-rs#44 for more info, but since this isn't technically in our spec (and I find it wasteful in any case to enforce), I think removing it is reasonable. I've already added a note to revisit this in #2060.

We could in theory warn users about the mismatch, but that would mean that everything produced by rustar-alginer (without a fix there enforcing the duplication of the index across multi-dim mappings) would yield the warning.

I'm open to suggestions, but this seemed the least invasive!

• Closes #
• Tests added

• Release note not necessary because:

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.84%. Comparing base (ad6c4f4) to head (0e8b204).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2641      +/-   ##
==========================================
- Coverage   87.60%   84.84%   -2.76%     
==========================================
  Files          49       49              
  Lines        8123     8110      -13     
==========================================
- Hits         7116     6881     -235     
- Misses       1007     1229     +222     

Files with missing lines	Coverage Δ
src/anndata/_core/aligned_mapping.py	94.91% <ø> (+0.38%)	⬆️

... and 13 files with indirect coverage changes