Class Balance and DPS Perception

There was an interesting quote today by Ghostcrawler:

"In the case of Shadow priests, we thought damage was a little low and made a very small adjustment. Many Shadow priests think their dps is still unacceptably low, but we don't agree. (That also doesn't mean we won't make changes in the future. We do change our minds.)"

The fascinating part to me is the disconnect between players and Blizzard -  and more specifically why that disconnect exists.  I think there's a couple of big reasons:

1) Sites Like WMO (wowmeteronline.com)
wowmeteronline is cool because it has top-20 lists for dps.  The problem is when players take top dps (or average of top 50, etc), rank classes by those averages and consider that to be represenative.  It's an inaccurate measure of class performance for multiple reasons:

• Randomness.  Take two hypothetical classes, Tortoises and Hares.  Tortoises do 5k dps on average +/- 500 based on RNG.  Hares do 4.9k dps on average +/- 1.5k based on RNG.  Now even though Tortoises do more dps on average, a site like WMO would show the Hares being significantly ahead - because all of the parses where a Hare gets bad RNG are ignored, and only the positive RNG ones are selected.
• Cooldowns/Fight Length.  Moonkins, for example, have Force of Nature.  Mine generally do roughly 40k damage, and I can use them once per 3 minutes.  40k/180 = 222 dps - but keep in mind that's a minimum.  If you graphed the dps gain from FoN compared to time, you would see 222 dps at every multiplier of 3 minutes - 3, 6, 9, etc.  And you would see spikes where the fight let you just fit in an extra FoN (30 second fights being the biggest, then 3:30, then 6:30, and so on).  Specs with cooldowns will therefore show as slightly higher on a site like WMO, because the parses where their cooldowns line up are going to be higher dps.
• Fight Mechanics.  Generally people ignore cases like Hodir which obviously favor casters.  But even beyond that, certain fights favor certain class/specs just based on mechanics, movement, or even strategy.  Consider XT, where the heart phase favors cooldowns even more so then other fights.
• Heroism.  Classes all scale a bit differently with heroism.  As a fight gets shorter and shorter, the classes that scale better will improve compared to others.

2) Self-Reinforcement
I was a feral (mostly tank, but I had dps gear) back in MH/BT.  Now feral dps was certainly sub-par because it just didn't scale.  But even if you took that into account, feral dps was lower then it should have been for a couple of reasons:

• Feral dps didn't get gear.  Loot council wasn't going to give it to them over other classes, and many progression ferals in a dkp system passed loot to rogues/hunters/etc because hey - they get more use from it.
• There's less incentive to min-max.  If no one expects you to put out awesome numbers, why does it matter if you do 1400 or 1300?
  

If there is a perception that your class does low dps, then that perception will create a negative reinforcement that will lower your dps even further - you will be less likely to get gear, to take gear (I know at least one of the spriests in my guild passes on certain upgrades because he feels other classes get more use out of them), and there's less personal incentive to do well.

And the real catch-22 of this is that these additional factors will reinforce that original perception of your spec, helping to cement it.  Perception that shadow priests do bad dps (maybe true) = shadow priests get less gear/min-max less = shadow priests do bad dps (true) = perception that shadow priest dps is low.

Caveat:
I don't know if ele shaman/spriest dps is fine now.  I'm doubtful that ~50 spell power would fix the spriest issues.  But I think the disparity players see is probably exaggerated compared to reality and the truth is somewhere in-between.