Clean dataframe math

[asmirnov82]

1. Remove redundant code from PrimitiveDataFrameColumn.BinaryOperations and PrimitiveDataFrameColumnArithmetic to make it more readable
2. Remove extra cast to slightly improve performance in PrimitiveDataFrameColumn.BinaryOperations
3. Remove converting to mutable buffers (involves memory coping) for operations where it's not necessary to increase performance in PrimitiveDataFrameColumnArithmetic

This is preliminary work for #6110

[codecov]

Codecov Report

Merging #6709 (1ba1342) into main (f93ab25) will increase coverage by 0.07%.
The diff coverage is 55.10%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6709      +/-   ##
==========================================
+ Coverage   68.77%   68.85%   +0.07%     
==========================================
  Files        1215     1215              
  Lines      251729   250697    -1032     
  Branches    26250    26256       +6     
==========================================
- Hits       173136   172607     -529     
+ Misses      71777    71276     -501     
+ Partials     6816     6814       -2     

Flag	Coverage Δ
Debug	68.85% <55.10%> (+0.07%)	⬆️
production	63.35% <55.10%> (+0.06%)	⬆️
test	88.88% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...ata.Analysis/PrimitiveDataFrameColumnArithmetic.cs	48.46% <ø> (-0.53%)	⬇️
...lysis/PrimitiveDataFrameColumn.BinaryOperations.cs	41.46% <55.10%> (ø)

... and 8 files with indirect coverage changes

[michaelgsharp]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[michaelgsharp]

I think a lot of this will be fixed by the GenericMath rewrite that @JakeRadMSFT is doing, but since we don't know when that will end up going in I'm going to go ahead and get this in for now so we can get the benefit of it asap.