my perspective is that the most significant percieved and discussed weakness of a unipivot is actually it's biggest advantage, which is the freedom to wiggle. it is the micro and nano wiggling following the groove unimpeded that gives it the advantage over a fixed/gimbaled bearing pivoted arm which on the micro and nano level cannot follow the groove as well.
A tone arm should not really follow the groove- it is the cartridge that must do that. The arm must keep the cartridge in locus. If you think about it, if the arm *really did* follow the grooves the cartridge could not make any sound! Thus we come to the idea of effective mass the the ever-important issue of mechanical resonance.
Unipivots however do not rule the roost by any means when it comes to freedom in the motion of the bearings. For example, the Triplanar arm employs an ultra-hard bearing that is so hard and so precise that Triplanar got investigated by the Department of Homeland Security because they were using more of these bearings than Boeing was. Triplanar maintains that a problem with all needle-and-cup bearings is that the bearings get damaged after only a small amount of use- whether gimbaled or unipivot design. That is why the bearings they use are so hard- 7 or 8 grades harder than the bearings in an SME 5.
The Triplanar has an adjustable azimuth system consisting of a worm gear that can tilt the arm tube. If you look at how a cutting lathe is built, its obvious that the azimuth of the cutterhead never changes- the cutter assembly rides on a extremely precise machined stainless set of tracks with stainless wheels. As far as I have seen, (and microscopic motion being the nature of LP reproduction) only a gimbaled arm can have the same kind of azimuth accuracy.