Kaufman once wrote the following about doubled pawns:
Now, the article outright states that doubled pawns are bad (about 1/8 of a pawn, but varying between 1/16 and a whopping 3/8), but it gives a number of caveats with that. I think it's consistent to read it as follows: backward pawns are bad in the sense that at least one of them often has some pawn structure defect (say, isolated, backward, crippled). If you already score the pawn structure defect (which is the correct thing to do anyway), then penalising the doubled pawn itself is redundant. Pu another way, doubled pawns are weak not because they are doubled, but because they exhibit some other defect. Which is really what I thought anyway, but I thought I'd try to back that up somehow.most doubled pawns (other than those arising from a capture by a rook’s pawn) are a liability, but to widely varying degrees. If they don’t fall into any of the above bad categories, and if the resultant half-open file can be occupied by a rook, then and only then can it be said that no other compensation is required for the doubling.
So it's possible that if a doubled pawn term helps in a particular engine, it's an indication that the other pawn terms are not optimally tuned. I have no proof of that though, that's just speculation.