Except it won't be 10fps 0ff. It will be about 0.9 fps off. The reason is statistical. For 10 fps to come off, the direction of the velocity errors and their magnitudes have to be simultaneously the same every time you fire a shot. The odds against that happening randomly are astronomical. The two are independent variables, and therefore their standard deviations are independent, so you have to add the effects of their standard deviations, which, it turns out, add up as the square root of the sum of their squares.
√((54 fps)²+(10 fps)²) = 54.918 fps average velocity difference for the combined error sources.
Now, if you get the cases perfectly prepped and matching in volume, and get them perfectly primed with weighed primers so that they don't become a source of error, and have all the powder charges matching to the nth degree, then your bullet weight influence will become the dominant term, and then you should be able to improve matters by weighing the bullets.
Note that none of this is to say weighing bullets won't help accuracy, even if they don't influence velocity variation as much as you'd hoped. This is because you narrow the tooling, so they may be ballistically more uniform as to BC and, therefore, more uniformly deflected by wind. If they line up in nice columns, check each column in separate loads to see if one shoots better than the others. If it does, sort more for that weight and keep them aside for accuracy applications.