Remove resample argument from IBMA estimators

[JulioAPeraza]

Related to #817.

Currently if resample=False we get an error if at least one image has shape/affines different from the mask. In this case, users will have to run an IBMA again with resample=True even if most images don't need to be resampled.

Changes proposed in this pull request:

• In this PR we remove that burden from users and perform the resampling only if the shape/affines are different.
• Also, we are removing the memory_limit argument, given that use_memmap decorator is not being used by IBMA estimators.

[codecov]

Codecov Report

Patch coverage: 100.00% and project coverage change: -0.02% ⚠️

Comparison is base (e59a083) 88.68% compared to head (1843993) 88.67%.
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #823      +/-   ##
==========================================
- Coverage   88.68%   88.67%   -0.02%     
==========================================
  Files          47       47              
  Lines        6026     6020       -6     
==========================================
- Hits         5344     5338       -6     
  Misses        682      682              

Files Changed	Coverage Δ
nimare/meta/ibma.py	99.57% <100.00%> (-0.02%)	⬇️

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jdkent]

LGTM!