If the question is if MA can survive with 10% returns then, my best understanding is this:
The amount deposited (thus the 10% of that) on the reports is not the correct number to go off of for MA income.
The key is the yearly turnover numbers, which are in the hundreds of millions usd per year.(167 m per year as of 2008ish)
There was a debate a long time ago in these forums about the meaning of turnover, and it turns out there is a difference in definition between Swedish and US accounting ideals. Which led to a general consensus, that it meant the total amount of PED that was recycled in one year.
Its not the deposits they take the cuts from, but the turnover of all the PEDs in game for (mostly) every time it changes hands/activities are made.
Turnover generally grows in time, except in the cases where withdrawals may reduce the overall pool, or the economy is seeing a large decrease in transactions/hunt/mine/craft runs, but it still turns over and over and over with every purchase and decay regardless.
One more thing to consider is that the return rate of 90% often neglects things like skill values, which are a factor as well, thus making the unseen total returns generally a little higher. Falling more towards the line with the MA goal/statement of x ped per hour cost.
This is why deed payouts vary. The turnover varies based on overall activity.
Even on minimal server activity the auction fees and decay per day are probably pretty sizable.
I would venture to say that MA probably makes a lot more in a micro-transaction model then say most of the current popular MMOs who are going to that Free-To-Play micro-transaction model.
Facebook and gaming trends show us micro-transactions are huge business. And MA has been doing it for a long time.
If I recall they completely paid the investment/startup costs (mainly to Jan) in 3 years. Which is outstanding for any business model, especially a niche concept such as this.
Deed sales, ship sales ,Land Areas, PP deals, ect. are additional capital they get to pay the salaries and development on top of that turnover cut, and shouldn't be ruled out of income for MA, as it is a sizable number as well on a yearly basis. It would make sense to pay the developer costs in the "sales" of new areas. Either by establishing a profit share until paid, or cash payments from sales. Afterall, its virtual, the only cost in the manufacturing of a land area is the developer costs (which are fairly minimal with a working and established framework to emulate, a good engine kit like crysis, and a good team).
I wouldn't worry about MA's income much. As it is, players will be making approx 400-500k usd per year in land deeds alone, and that is just a fraction of calypso revenue.
I also think those 60 employees, don't have much to worry about either. (unless they split off to another company, but that is...something else)
at least, that is my understanding.....