When the politics are taken out of it, it's pretty simple.
Since NGC began their Left Alignment paper inserts (rather than the centered alignment inserts) they have gotten strict to the point of being grumpy.
There are some coins, such as
ASE and most gold, where NGC has many times the number of coins graded as does PCGS. So you would think their would be a lot more NGC on the market, too.
Yet no matter where you look, there is a lot more PCGS out there for sale, in almost any category (with a few type exceptions and
ASE's and very recent commemorative s).
For those poor souls who have decided to stick with the NGC registry, finding the slab is a lot harder than finding the coin. Which is the exact opposite of what we are taught to do.
Certain areas where PCGS has become noted for their lack of strict grading, an example of which would be the common date and varieties of the Bust Half Dollars. NGC has gone the opposite direction and actually become stricter. You can hear the talk at almost any show of any size.
So right now, the result is the exact opposite of gradeflation, which everyone likes to yap about. For some areas of the NGC encapsulation, the results are making it bad for folks who open their returned orders only to find out they were a grade, or even two, lower than they expected from NGC.
This will have the effect of a lot of submitter's returning to PCGS, just to get a better grade, which will further compound the problem.
I've been snatching recently graded NGC stuff as fast as I can find when I can find the coins I like. I still like buying coins, not plastic. But I will say that when I've put a few of my new NGC slabs in next to a few older PCGS holders and examined how both look one against another that the assigned grades are not consistent. New NGCs given AU-50 compared to a PCGS of the same grade a year earlier or later, and you can really see a difference in a lot of cases (for the Bust Halves, again).
Learn to grade and it's all irrelevant. Then the only person you have to worry about is yourself.