Much of the paper is exactly devoted to how they came up with the time scale. There are regular rhythmic deposits in the cores they are looking at. The authors rule out regular changes in orbital cycles that would alter insolation patterns as being too long. The only other possibility would seem to be annual cycles.
As - as far as my limiting understanding permits - it seems fairly robust to me. That doesn't alter the fact that it's logical to look for the weakest link here? Once you've ruled out the impossible, whatever is left - however improbable... and all that?
If someone can come up with cycles likely to occur between these extremes that could possibly result in the layering they describe, that could go a long way toward explaining the apparent instantaneousness of the changes they describe and would be most welcome. But just saying one doesn't like the time scales and pontificating about the scientific method is not much help to the discussion, imho.
I don't like the timescales simply because it's so tremendously difficult to comprehend mechanisms that can make them work - and I'm not sure if they fit other information about the PETM extinction (that is much harder to quantify of course, and I can't even do an adequate job on the mechanisms).
Incidentally, another objection I can think of about the cometary carbon idea - to counter anyone who thinks you can conveniently park something that big in an ocean:
http://en.wikipedia.org/wiki/Impact_depthThe question really is - does something that big moving at that speed act as a Newtonian impactor - or do other factors come into play?
If it does, the implication is clear - the object was minimum 13km in diameter (if pure carbon, and not impacting on land where you'd expect a big signature) - and I'm not sure where there's enough water to park it without still leaving a crater (or at least a mark)? One can play with the angle of incidence - but the more you decrease that the lower probability the event becomes.
I'd also be extremely skeptical of something that large blowing up into little pieces in the atmosphere (eg Tunguska) - the Wikipedia article above suggests the atmosphere is good for around 10m of water - against 13,000m+ of impactor?
If it arrived as lots of little pieces, it's hard to see how they'd all have been small enough to hide and still densely packed enough (in time and space) to have that level of effect.
Likewise if those few million years were characterised by multiple hyperthermals - I would've thought release of methane via volcanic activity rather fluky and likely to be a one off?
With respect to submarine clathrates - am I right to think the planet was already pretty warm and ice free at the start of this process? If so - that further undermines that argument (at least taking my view that only shallow water clathrates are likely to respond especially abruptly) as shallow water clathrates are necessarily stabilised by temperature and not pressure - and hence you require areas with the Arctic climate to retain them. While I still think shallow clathrates in the Arctic are a real threat for us today, I'm currently less and less convinced they fit the scenario described by the paper.
The challenge for anyone is to find a way to justify and explain their position - whichever possibility you favour. Personally though, it feels like something way beyond my knowledge/understanding to think of a good answer - even though it keeps nagging me as I hate unsolved problems.
Still, I've learned quite a bit trying - I stumbled over something interesting about the formation of coal that suggested so much was formed during a particular period in earth history as the organisms to digest it hadn't got going (more on the coal thread).
So there are certainly plenty of things about the earth system most of us don't know, and plenty more that nobody at all knows.
Maybe we're looking for some other unexpected quirk that could explain the hyperthermals and the speed of change - something unique to those few millions of years of earth history?