Machine Learning vs RFM/RFV

I'm looking over Dataro’s suggestions and there are some surprising inclusions. Are they correct?

When it comes to identifying potential donors to contact, machine learning will usually identify some donors that were not identified using more traditional models like RFM or RFV. This is a key benefit of using more advanced propensity modelling approaches.

Unlike RFM or RFV, which only consider three features (the Recency, Frequency, and Monetary Value of that donor's gifts), Dataro's machine learning models approach take into account hundreds of different model features covering giving history, demographics, interactions, etc. By considering a wider range of features, Dataro builds up a much more nuanced view of each donor's propensity to give, allowing fundraisers to make more informed decisions.