If your organization has a good training program you should have a ZERO or near ZERO check ride fail rate. Near ZERO allows for the guys who turn into a bundle of nerves on check rides, or the check pilots with an axe to grind.
If your company has no continual training program then you probably have a pretty small fail rate with internal check pilots, but if an outside source accomplished the checking it may or may not be higher.
Pilots who can not attain and maintain competancy should be mostly weeded out during initial training.
If the company has a good training program, the check rides should only be finding points that need additional emphasis in training, not really things to take a pilot out of the seat. That should happen well before he is to an evaluation event, either initial or recurrent.
You will always have the occasional dirt-bag that gets thru, but they should be few and far between. Usually these guys are brought to the company's attention by co-workers or customers before a check ride happens.
There are always exceptions to these points, but generally this is the world as I see it. The key to success is the training program, initial and recurrent. This recurrnet program should not be a short burst once a year either
----- Now how do we get training budgets and assets available?
HTAWS, GPS, NVG are all tangible and you can point to it as a step you took. Training is not so much easy to portray that way.