Consistent notation for learner APIs

[EgorKraevTransferwise]

Would it be possible to harmonize the notation across the different learners? For example, the treatment propensity model is referred to as model_t in CausalForest, model_T in DMLOrthoForest, propensity_model in XLearner, and as model_propensity in ForestDRLearner. Same for model_y, etc

That introduces confusion and complicates code reuse, for no good reason that I can see.

[kbattocchi]

This is a good suggestion - thanks. I don't see any reason not to harmonize these.