It is very difficult to prove today that a painting done two or three hundred years ago, especially one without a signature or with a questionably authentic signature, is indubitably the work of this or that particular artist.
Background
You can't really prove (from analyzing the style / techniques / pigments / etc) that a painting was done by Person X vs. Person Y.
So if there is an attribution already attached to a painting, it is given special weight, since we assume that the attribution has been consistently carried down from when it was first added 2 or 3 centuries ago to today.
Position
Traditional attributions shouldn't have special weight.
Evaluate / Goal
We're just strengthening a position, not an argument, so we can really just head to the answers and think,