[bugfix] Correctly multiply weights for FastDG, when using accumulate
In sum factorization code, we are usually doing the accumulation ourselves, so this one went unnoticed.
In sum factorization code, we are usually doing the accumulation ourselves, so this one went unnoticed.