I don't have those unbiased studies, sorry. I do remember reading an article on how a journal bearing turbo can handle, for instance, a boot blowing off better than a ball bearing. It was explained that the side loading from the imbalance in pressure turbine to compressor can overcome the ball bearings strength. This typically breaks shafts also... I don't know how bearings fail specifically, but read they do. I of course am not an expert, just have done some reading. Take that for what it's worth. Oh, and all types are "floating on oil." Engines wouldn't last very long otherwise.
On oil, I also have read that with ball bearing turbos, cooling EGTs before shut down is more important. Coking of oil preventing the balls from rotating comes to mind, but I can't be sure this is the case. I also can say I don't know of 38R issues from this. Just a talking point I guess...
The point of the ball bearings has always been to reduce friction and aid in spool up. The wheels have progressed to the point that the ball bearing turbos from 10 years ago won't spool as fast or make as much power as current up to date journal bearing. For a daily or a tow rig, I would venture to say a properly sized journal bearing turbo will outperform a ball bearing with a mismatched housing all day. For instance if a .84 is optimal, and a 1.0 is installed. Speculation yes...
My understanding is this, bearings have 3 jobs. Keep the turbine centered, keep the compressor centered, and control thrust (axial shaft movement). Ball bearing turbos aren't necessarily ball bearings controlling all 3. Some only have 1, and the rest is same as a journal.
My suggestion @tomlin is to call and talk to some turbo builders. They are the ones that are seeing and dealing with failures daily. They will be a good source, and if they all are saying roughly the same things, I'd take that for truth. I do get wanting to have real evidence before putting your stamp on something to a customer.