...
However note that the rate of growth has started to slow down instead of speed up at around November '13:
That's actually kinda good. The sooner we get to a mature mining market, the better for the security of bitcoin in general. I'll consider bitcoin a little less exposed when a few things happen:
1) There are no more hash/watt exponential gains to be had from a mining-tech jump; eg, going from 28nm ASICs to 22nm, etc. We may at that point now, actually. The gains from new process tech are certainly diminishing (as the taper in difficulty rise probably shows).
2) The power requirements of a 51% are an order of magnitude bigger than the world's largest datacenters.
I also look forward to a 'saturated' mining market. Miners would operate at marginal cost and a very large part of expenses would be operational (power, maintenance), not capital investment. Everything else being equal (especially bitcoin price), hashrate should then grow according to moore's law and miners would sell close to all coins into markets.
Last year I've been thinking we might reach your point 1.) in the first half of 2014. This hasn't materialized, mining technology is still improving faster than moore (still catching up on other fronts).
I'm not sure your point 2.) means less danger for bitcoin. A 51% attack could be distributed across multiple datacenters or done using pools. It's hard for me to imagine someone (govt?) doing a 51% attack this way anyway. The most likely scenario I think would be governments forcing miners (51% of them) to censor transactions and consider blocks containing censored transactions as invalid. But for that to be a meaningful tool for them (say they want to cut off wikileaks for example) they first need to have better coin-tracking (hello coinbase, circle,...)
