The Steppi performs better ;o)
Just have a look at the pictures in this german thread:
PCBMillingTable
First image shows the prototype of a PCB based heatbed for a 3D printer. What you can see is, that, given the PCB itself is evenly thick, a PCB-milling table helps to have even-thick traces across the whole PCB.
Next you see my test for SMD. You can even have a trace between 2 legs of the SMD.
The next images show the bits I use. Not cheap, but very good. And with hard-paper-PCBs they also last for a while.
#12277 shows the measurements of the traces.
At the moment I try milling PCBs with EstlCAM software + Arduino, which allows to probe the surface before milling.
I'd really like to build my own PCB milling spindle running at something around 60000 rpm using a brushless motor and maybe extend the PCB milling table by an oil-bath for cooling, better cuts and extended lifetime of the bit.
End goal looks similar to this:
PCB
(With end goal I only mean the look they showed. I think the Steppi is better than the wegstr, as they also don't have traces between legs of the SMDs and it looks like the Steppi is milling faster than the wegstr)
I already tried to build my own "spring clamp" for scraping off the soldermask - so far no luck because I don't have a lathe. I drilled a long hole into some 8mm aluminum round rod, glued in the stump of a killed bit on one side. But this way I ended being off for 0.1mm from the center. So, mounted to the motor I have too much vibrations. Experiments will go on ;o)