Back on the torque - I dont think that grease on the threads has much to do with it, in other cases (e.g. cylinder heads) you can get misleading torque values if the metal binds rather than sliding, so greasing the threads actually helps on a longer bolt to ensure you are not just twisting the shaft of the bolt with the torque wrench, only for it to pop back again. I have used 28 ft lb torque, Schnorr washers, spreaders, 12.9 hex socket cap bolts off UK eBay suppliers, driven 90k miles , they always seem to stay done up - the slightly softer spreaders are critical because if they move around the bolts stay locked by the Schnorr washers onto the spreader plates. Without them, the washers will start to wear down against the harder metal of the CV joint outer shell.
Yeah, you may be right about the impact of grease, but then again There's no doubt that lube prevents galling on dry threads but how much and when? I don't know what the spec is for preparing the CV bolts, but apologies if I gave the impression I advocate dry assembly. What I should have said was that excessive grease between the threads, and the flange/joint face may play a part in failure. Less is more on these areas. A note re your 28ftlb. That's 37Nm. Tighter than the widely quoted T2 spec, but your bolts and joints will handle 45Nm or 33ftlbs if you wanted. That's the spec for the rear of the T3 which uses the same parts. Interestingly the front inner CVs on the syncro are specified at 35Nm...go figure. Less torque transfer I guess.