More variation public doc updates

[nakib103]

Some more updates to public docs for variation -

• Revert back protein function pages mentioning we are using 3 bits for prediction and add a note it is only for AlphaMissense and ESM1b.
• Update filter_vep docs mis-using is operator for SIFT/PolyPhen score when both prediction and score is present.
• Fix rows for rabbit and japanese quail in species_data_types.html