If I follow the approach procedures I will stay away from the
granite. I would recommend reading TERPS.

I'm certain that Airperson is far more familiar with TERPS than you
are or ever will be.

His point is that certain types of procedures lend themselves to
certain types of errors. Nonprecision approaches have a far higher
error rate than precision approaches. The weak link is the limitation
on human ability to manage complexity; the only solution is reducing
complexity, either through technology (which is what a glide slope
does) or via pilot training, such as the constant rate descent.

To comment that "if they fly the approach as published, they won't hit
anything" is a very shallow analysis. Not every CFIT deviated
intentionally from the published approach.