Yeah, it's like a greater percentage of balls hit to RF will be outs in a pitcher's park, but a few more will drop with a poor range RFer.
Arbitrary, crude, totally made up number example:
Say, 50% of balls would typically be outs in a neutral park. A pitcher's park might be more like 70%, hitter's park more like 30%.
A catcher in RF would bring those numbers down, say, 5-10%.
The effect in a pitcher's park would still be net better (fewer hits) than a neutral park, but in the hitter's park it would start to get out of hand.
At least that's how I figure it works....