During previous mass extinction events, carbon isotope excursions consistent with the release of hundreds of billions or trillions of tons of carbon from the methane hydrates occurred. This has happened several times, not just once or twice.
Generally there is flood basalt activity going on, releasing massive amounts of CO2, much like our current fossil fuel use. But the flood basalt activity generally seems to precede the sharp mass extinction event, sometimes by hundreds of thousands of years, and often seems to go on after the mass extinction event.
Methane release from the hydrates, triggered by flood basalt release of CO2, is the best mass extinction hypothesis to explain all of the geological evidence, I think. The sudden methane release hypothesis has the most explanatory power, the most predictive ability, and even makes quantitative predictions that turn out to be correct. It is consistent with all the geological evidence. When a new claim is made that contradicts the methane release hypothesis, that claim generally does not hold up, and turns out to be wrong.
The methane release hypothesis turns out to be a unifying theory – it constitutes a general theory of most or maybe all mass extinctions.
Link: http://robertscribbler.com/2015/09/10/new-study-risk-of-significant-methane-release-from-east-siberian-arctic-shelf-still-growing/#comment-51528. Nota bene an excellent graphic of the End-Triassic with C-13 content spike consistent with the End-Triassic mass extinction.
Leland Palmer further notes that we don't want an End-Anthropocene extinction either, but with all the fossil fuels we've combusted up so far we just might get it. With 6,800 billion tons of methane clathrates in the Arctic and 1,600 billion tons of Carbon in the Arctic permafrost, I agree.
UPDATE: Mr. Palmer has since posted a graphic of the negative Carbon Isotope Excursion (CIE) from the End Permian -- the mother of all smelly hothouse extinctions.
These repeated sudden CIE events, coupled with hyperthermal mass extinction events, are good consistent evidence of methane release from the hydrates. There have been maybe 20 or more of these flood basalt eruption / sudden negative CIE / extinction events, and hundreds of smaller apparent releases, I think. So, maybe our hydrates are less stable than we think.
Considering that the hydrates up in the East Siberian Arctic Shelf are already letting loose, I think they are a LOT less stable!