Teevee,
What would you accept as evidence?
The best place to look would be in a broad sampling of FOQA data spread across carriers, aircraft types, nations and cultures. Look for exceedances that relate either to hand flying or where poor monitoring allowed the autoflight to create an exceedance. Then look at the relative rates of manual approaches. Do that and run a rolling 5 year trend over a period of say 10 years and you'd probably get a pretty good idea of whether you have an increasing problem. pj2 is the man who could tell you best how to do it.
If you didn't have the FOQA data the next step up would be to look at cumulative incident reports. Last, and least effective is to just look at accidents. It's the precursors that never show up on the 9 o'clock news that are the best indicators of whether there's a trend that needs correcting.
ELAC