Having a nuclear reactor onboard is the opposite of a problem in a case like this.
Clearly you'd rather a collision between two 1kg satellites than two 1000kg ones - where does this logic break down as we make one or both sats bigger?
Satellites don't have this constraint. Of course you'd generally prefer smaller sats to collide (smaller = less fragments), but assuming fixed size, denser (= heavier) pieces will have more inertia, and so deviate less from their original flight path. This keeps the debris trail more concentrated. If only one of the two is heavy, the pieces of the heavier one should end up more concentrated. I'm not sure whether the wreckage of the lighter one gets spread more widely, or - with these velocities - just flash-fuses itself with the heavier one.
Taken to the extreme I'd rather hit a feather than a bowling ball at 14 km/s.
But what matters here isn't what you would rather do, it's what both you and the feather/bowling ball would rather do, in context of minimizing the size of the resulting fragmentation.
Intuition tells me that a hypervelocity bowling ball will barely notice you, much less have time to deposit lots of energy into you.
The bowling ball collision has a lot more energy involved, it can eject more debris at high velocities in various directions.
The heavier the fragments and the less eccentric the orbit of untracked space junk, the more dangerous it is, because the longer it'll be hanging around before it gets close enough to Earth to de-orbit itself.