I once worked in a company that used the LPKF mill to drill their prototypes. They didn't use it as it was advertised because (a) it broke far too many mills, about 10 per board, in which case it ruined the board, too, and had to be restarted. (b) It "smeared" the copper about 30% of the time, which left copper where it shouldn't have been, therefore requiring a restart.

Overall, the board yield was so low that it was quickly abandoned. Aside from the physical limitations, e.g. track width, hole sizes, dimensional tolerance, it was just too much trouble, producing a useable board only about one time in ten attempts. It wasn't very quick either. That's the reason they now have a state-of-the-art PCB fab. Now they can produce 16 layer boards, hold tolerances well below 1 mil, drill holes as small as 2.5 mils, register layers with 5-mil inner layer pads and reliably place evenly spaced tracks as small as 3 mils. I'm told that it costs less per board than using the LPKF board mill. They seldom produce more than three or four boards in their prototype lab.

RE