I won't bother discussing this in depth since no matter what I will say it won't change your opinion, if for no reason then just because developers claim that it didn't change a thing. The fact that they used a single instance and not particularly attractive outside of the lore as the source for that data no one cares about. I assure you if it was Warrior that was added while Dark Knight and Paladin would have been already in the game, the results of that "study" would have been different.
However, you're asking about players starting to play tanks over DPS. But I'm talking about the other way around. There ARE players that moved AWAY from tanking and healing TO DPS with Samurai and Red Mage. And as DPS are typically more attractive to players in general outside of the queue times, these players that move are much more likely to stay a DPS unlike those that try a new tank or healer.
It doesn't matter how many people you pull in into tanking and healing if these players won't cover for those that STOP tanking and healing. Either moving to DPS or quitting cause they're bored having only 3 classes they might like (and in healers case, three very similar classes to boot) when DPS have nine and prospects are for DPS being catered to while tanks/healers not.
And how many times did that happen?! Once. Do you know what qualifies as research?! Doing the same thing over and over again and comparing all the results.
Things that are based on players preferences, moods and personal availability/situation are muddled up even more. You say that the queues are the same as they were in heavensward around the same time, but why?! You don't know. And I'm sure that the developers don't pry deep enough either.
There are many reasons why such a thing could have occurred. More DPS players leaving the game than tanks/healers. More of the tanking/healing group players buying the game and starting to play. More DPS players taking a break from the game than for tanks and healers.
Since there are more DPS players than tanks AND healers combined, all those "more DPS" is statistically the scenario for whatever we're discussing. If 10% of every jobs mains would take a break, far more DPS would go on a break than tanks and healers combined, not only because it's 10% out of nine classes instead of six, but more importantly because those nine classes in total have more individual people than tanks and healers combined.
For all we know, if there would be a tank and a healer, the queues could have been BETTER at this time. Hence there is no sufficient data to amount to anything more than hearsay or speculations. And it's the same for developers, except their guesses are more educated and have a higher chance of being correct. It's just that there are too many variables to accurately pin-point a single cause as the major 'player'.
...And for that, we need new tanks...you know. People have been complaining about tanks and healers having their core kit pretty much copy/pasted since I started playing this game about fourteen months ago. And there were added balance issues on top of that. Square Enix needs to get out of the "comfort zone" and put some work into variety among tanks and healers. But they follow the yells of the DPS crowds and spend more time creating engaging DPS classes while tanks and healers are an afterthought. And you expect the population to be even remotely balanced?! What classes changed gameplay-wise with stormblood the most?! DPS. Summoner, bard, machinist and black mage are completely different from what they were. And tanks?!...Yeah. They just lost tools and had potency changes, and that's the biggest change. Of course, there came new skills, but those are a given. Then there are the latched-on gauges, which really, matter only for a warrior (the most DPS'y of the tanks...). They feel forced on dark knight and paladin, and feel more like just another form of cooldown than anything.