If maintaining a minimum of 50BTC per block is the goal, then we're going to need to see much higher volume of transaction. I'm not sure this is the goal. Perhaps in 20 years, when BTC is widely used and blocks are worth 1BTC, transaction fees will only total 10BTC. I think this is also an acceptable result.
Block mining payouts of 11 BTC will only be acceptable if 1) its profitable using the hardware required and 2) Its cheaper to mine BTC than to buy it.
Both of these depend on the fiat currency value of BTC at that time (or the barter value!); a completely unknown quantity.
If the 50BTC per block payout is not maintained next year, I personally suspect mining except for the hardcore miners will collapse. Miners will have to be able to produce multiple GHashes to be profitable (prob already so). In the meantime I'll be buying my BTC I think, and possibly doing a little FPGA mining only if its cheaper than buying from the exchanges.
With these assumptions, I reckon next year will be crunch time for BTC!