Finished looking at the thread... a few opinions and speculations follows.
Gear iLevel beats reforging
I see so often people underestimating iLevel and what a tremendous advantage this gives a person over other players. This is a gear-based game. When I look at the priest singled out in that MMO thread, what I see is a priest who is drowning in heroic gear while the rest of the world is nowhere near as geared as him. This guy can do a few things wrong and still get most everything else right, and still come out looking like the "best priest ever".
My theory about people like this is the old addage - "Using a bazooka to kill a fly." This person's gear outranks everyone by such a wide margin that he's going to demolish anything he aims at. We can't really be sure if there's true genius in reforging at work here, or just over-compensating for small issues.
Sims don't match
Nowhere in any sims I've seen show mastery approach crit. This leaves me with two possible conclusions.
1. The sims are not accurately modeling shadowpriests.
2. The sims are correct, and this shadowpriest would have done even more dps if he had been reforged properly.
It's really that simple.
My experience thus far
My experience with multi-dotting so far has given me the following experiences.
FDCL past about 3 targets makes you hit the GCD cap. You hit a point where VT/SW: P starts falling off, and it becomes more important to refresh them and/or cast MB/DP3 than to cast MS. At the same time, FDCL doesn't proc when you want it to, so you end up with many wasted procs back to back before you can use them. Sometimes you're so busy casting other high priority spells, the proc runs out.
The process plays and feels very non-intuitive. As a result of this, I honestly don't feel mastery + multidotting + FDCL is some "magic secret" that we as a community need to keep pursuing.
I think what's more than likely is that there aren't enough priests at this high of level of play yet to offer us any new views. Additionally, I think priests at this level are probably not experimenting much, because they cannot afford to. They find something that works well and sticks with it because each boss attempt, certain results are expected. The raid can't afford several wipes while you mix and match talent specs trying to find which one does the most dps.
Unfortunately, a lot of us are playing in a bubble, and when you play in a bubble, you're always the most awesome spriest you know... it's very difficult to say whether what you are doing is "the best" when that's your only view.
I think we need more data on this. I especially would like to see some type of sim data backing up the idea of stacking mastery.