[TabArena-v0.X] Try better fillna / cat encoding strategies for RF/XT/KNN

Try better fillna / cat encoding strategies for RF/XT/KNN

Currently ordinal encoding for RF/XT and fillna 0 for RF/XT/KNN.
KNN ignores categorical features.

Figure out what missing categories are assigned to for RF/XT.

Maybe try mean fillna + indicator for numeric.

Start by tracing the RF, XT, and KNN preprocessing entry points to identify where ordinal encoding and fillna 0 are applied. Compare alternative categorical and numeric missing-value strategies in the TabArena benchmark, verify how missing categories are assigned for RF/XT, and document the resulting model comparisons.

Written by the indexing model from the issue text.