IDF is used to measure how important a word it is (common or rare). In other words, it is a measurement score of word popularity. TF-IDF can measure the document relevancy for a given word. Can we directly use it as the score of “Terms Match”?
For others, to complement the infos of your post:
And I think you are right. TF-IDF sounds like a good candidate tom compute the “terms match” part of the score
In the Document Selection process we want to select relevant documents that contain the query terms. Thus, we care about IDF.
If we use TF-IDF, then we are assigning a weight that signifies the importance of the term in a particular document.