One could make a reasonable argument that if their balance sheet was so tied up in sketchy mortgage assets that they were forced to meet margin calls when the market fell, then they weren't as solid as they seemed. Their problems started before they got in the car.
I realize that there's a bit of begging-the-question here, but even so, if you've borrowed a ton of money and secured it with "assets" of questionable value, then you're taking a risky bet. The accounting regulations didn't make the bet better or worse -- they just kicked the losers out of the game before they wanted to quit.
The purpose of mark-to-market exceptions are to allow temporary price volatility not to drastically impact a firm's balance sheet. The FASB rule specifies that it's an acceptable practice to suspend mark-to-market in "fire sale" conditions.
But the real issue is the flawed risk ratings on the MBSs. If your asset is AAA then it really shouldn't matter whether you are marking it to market or not, as it's supposed to be quite liquid, by definition, as there is supposed to be extremely little risk that its price will fall.
The reason the MBSs weren't liquid is because the ratings were incorrect and the market knew that -- perverse incentives from ratings agencies had led to this, as had the widespread belief (taken as an assumption by many foolish risk-managers and codified into pricing formulae) that housing prices could not fall.
If the MBSs had been properly rated, then mark-to-market would not have caused the firms holding them on their books to become insolvent, as capital adequacy requirements are based on the riskiness of the assets.
So while mark-to-market may have sped up the failure of those firms, the firms were already weakened due to their treatment (thanks to bad ratings) of risky debt as AAA.
In fact, a few months before things got really bad the SEC sent out a memo advising that the fire-sale provisions of the FASB rule regarding mark-to-market were not to be used, a decision which suggests either that the SEC was OK with some firms failing in the near future (unlikely) or that it believed that the ratings were generally accurate (even more unlikely). I've been told by some Wall Street experts that everybody knew that the ratings were BS. Hence the SEC's decision to send the memo was probably just intended not to telegraph the possibility of a future problem to avoid spooking the market in hopes that the problem would go away.
A few things to note: The regulatory trend is moving toward BASEL II which relies heavily on both mark-to-market accounting and third party ratings agency ratings. So the big regulatory question is, how can the perverse incentives that arose and led to the inaccurate ratings of MBSs be restrained? So far everyone is talking about banning the trading of x, y, and z and of limiting bonus pay, executive salaries, etc., but nobody is talking about the actual concrete perverse incentives (pressure on ratings firms to issue inaccurate ratings) that fed the whole mess.
I think a requirement for a random subset to be inspected by a real independent group like the SEC would cut down on a lot of this crap. You would see a lot of stuff get through but it would put a dampener effect on bubbles. You would also need to keep track of how the agency rated vs each company.