I'd just like to hear how furuknap believes mining can be profitable AND difficulty not rising - yet there's no significant interest in buying more mining gear. Because somehow that's a scenario he bases part of his case on - and it doesn't exist. Difficulty stops rising when mining is UNPROFITABLE if capital costs are taken into account (those with ASICs already continue mining as the capital cost is sunk and resale value becomes low).
I believe mining is designed to balance profitability in a way that makes it, over a long enough time, marginally profitable. Right now we are in a position where new mining operations are not profitable if the current situation continues. Thus, by the same argument you are making, difficulty will cease to rise until such a time where mining operations again are profitable.
However, like with mining investments now, this change isn't sudden. There is a significant lag in the market's perception of long-term mining viability and investments turning into real mining efforts. During this time, as we see now, mining is incredibly profitable thus driving future investments further up.
Even now, long before the full force of the previous months of mining investments have started operating, we see a rapidly rising skepticism of any mining investment (not just assets). Just look at how skeptical people are of PMB style assets or ASICMiner hardware and how slow the new Bitfury and KnC miners are selling.
Thus, when mining becomes barely profitable or even factually unprofitable, hardware prices may drop rapidly to make mining profitable again. This assumes that hardware vendors have large markups now or can continue to sell current generation hardware indefinitely to avoid NRE.
However, unless hardware vendors stockpile huge amounts of ASICs, we'll see the same lag in turning investments into mining operations, perhaps as long as we've been waiting on ASICs to make mining unprofitable now. This lag will cause mining to be very profitable for a long time before new hardware arrives.
In short: For every day of vastly unprofitable mining we'll see (or have seen) one day of vastly profitable mining. Over time, it is designed to balance out and the question is not if this will happen (because if it doesn't, Bitcoin is doomed) but when.
.b
Mining isn't "designed" with any sort of consideration of profitability at all - any more than gold is "designed" with profitability in mind or any other commodity is "designed" with profitability in mind. Mining exists - and profitability or otherwise depends on how it is used, not on any other aspect of its design.
Mining IS designed to produce fairly stable coin generation. That DOES impact on profitability - but isn't why it was designed in that fashion.
You say :
"Thus, by the same argument you are making, difficulty will cease to rise until such a time where mining operations again are profitable. "
No. At no stage have I ever argued that mining needs to be profitable. My belief, in fact, is that if technological advances didn't occur it would cease to ever be profitable at all (and would end up sat at a level of difficulty that didn't move much at all). That's because there ARE various classes of people who will mine even if doesn't make a profit:
1. Those doing it for fun or to support the bitcoin network, where profitability isn't a factor of importance to them.
2. Those with 'free' electricity - where although the mining isn't making a profit overall it IS making a profit for them.
3. Those who fail at math.
4. Those who buy mining hardware then sell shares in it to others who fail at math/common-sense.
Add that to those locked into hardware already - whose loss is already mostly fixed - and difficulty will happily stay above the level where mining is at all profitable.
Technological advances break that - and mean that, for short periods, mining is decently profitable (but with ASICs that decent profit isn't reaching the general public). And ongoing advances on a smaller scale (decreasing cost of unchanged hardware) can keep it marginally profitable.
And even if it's not profitable at all, you only need to find some suckers willing to pay 5,10 or 20 times the cost of hardware plus add a favourable buy-back clause for issuers to make a profit from mining (even though the profit doesn't actually come from mining - but from giving back investors a lot less than they invested).
And there's one HUGE way in which DMS.Mining isn't like PMBs/mining contracts (ones which pay out on PPS or similar). Even if mining becomes unprofitable DMS.Mining still has proven backing in place to meet its contract. If mining output falls such that cost of operation eats up a large chunk of mined coins then most PMBs have no established backing from which dividends will be paid - they just have to rely on the operator digging into his pocket for funds he may or may not have (history suggests quite a few won't have it or will be unwilling to part with it).