Why is the default n_trees=1000 in ForestDRLearner?

[EgorKraevTransferwise]

That seems rather large, especially as in CausalForestDML the default is n_estimators=100, and the ForestDRLearner with n_trees=1000 can take quite a while to fit. Also for example in XGBoost, also a tree method, the default number of trees is 100.