That is certainly true if we were looking at isolated runs and trying to make sweeping determinations based off of those results. Instead we are looking at a series of best-case runs within a demographic whose skill level is within margin of error of one another.
The logic here is as follows: With every possible combination of group composition, route, playstyle, and with an unlimited amount of attempts per run to achieve the best-case result, Guardian druids are sitting at a M+ wall of approximately +24. The same factors being true for Prot warrior, they are sitting a M+ wall of approximately +26.
Another way to look at it is by using IO score as an aggregate measure of ideal-case M+ performance. You may disagree with using IO score as an imperical measurement, and to an extent I agree that it is nowhere near perfect. That said, I still find it to be very useful as an indirect measurement of performance.
Looking at # of tanks with >2800 IO score:
- Prot Warrior: 921
- Blood DK: 288
- Prot Pally: 248
- Vengeance DH: 179
- Brewmaster Monk: 65
- Guardian Druid: 39
Likewise, widening our view to >2400 shows the same relative trend as above, albeit with Brewmaster and Guardian switching places this time around.
Looking at # of tanks with >2400 IO score:
- Prot Warrior: 9067
- Blood DK: 4313
- Prot Pally: 3167
- Vengeance DH: 3000
- Guardian Druid: 1390
- Brewmaster Monk: 984
I will note some MAJOR shortcomings with this approach. First being that these numbers are not proportional to the number of tanks within each class bracket. This means that we cannot say for certain whether or not there are more IO scores for a given class becasue a class is underperforming, or because there are simply more people playing that class because it is more fun, popular, easy to play, etc. And second being that these numbers conflate both pre-10.0.5 and post-10.0.5 totals into one, and the issue at hand is not only whether or not Guardian druid is weaker than the other classes, but more importantly if 10.0.5 made them worse. (I tried really hard to pull data for total # of characters and the date each run was recorded, but without learning their API and sobering up a tad, this is the best we’re going to get.)
So, take the above numbers with as large of a grain of salt as you think is necessary, though, in my opinion, these numbers are still pretty interesting nonetheless and shouldn’t be ignored. What we can say for certain is that Guardian is a particularly underplayed spec compaired to the other tanks specs, with the only exception being to that of Brewmaster monk. My prediction is that the addition of the major nerfs in 10.0.5 is likely only going to widen that divide even further over the coming weeks and into season 2.