I'm not sure that Moores law applies to the block propagation latency - it was argued that even 1MB is too big, as miners far away from the Chinese mining centers would be wasting significant time mining atop at an old block. Then again, with technology like Matts relay network, heads-first mining, xtreme thinblocks, FIBRE, etc, the argument is not relevant anymore.
The argument that the hardware requirements are too heavy for "ordinary people" to run their own nodes - at one hand, I can feel it myself as the equipment I had dedicated for this purpose doesn't work out anymore (I turned my node off - one of my servers didn't have enough memory, the other didn't have enough disk). Then again, everyone that really has a need for running their own node can afford setting one up - and I bet the total amount I've paid in mining fees amounts to several dedicated computers running Bitcoin Core or similar software.
I just realized that you never answered my question lol. I know that it is pure speculation but it is still fun; how high do you think this run will go? How long?
Come again, what was the question?
In Bitcoin context, Moores law is usually used to explain that network growth can be handled by increasing the block size, and that the cost of resources handling the network growth will stay constant due to Moores law. The usual counter-argument is that the biggest problem with bigger block is the latency involved in broadcasting new blocks. The bandwidth available is growing, but not at the rate predicted in Moores law. Hence my answer above.
Until the 1st of October - unless news appear in the scaling debate, the bitcoin price will not grow more than 5%. According to my own tracker, the current market rate is 27244 NOK/BTC, so at the 1st of October I believe it will be 28606 NOK/BTC. (I don't follow the USD rates, they are pretty much irrelevant for me).
Then again, I believe we'll see a lot of black swans before that time. Bitcoin price could collapse on bad news - it may be that we're in a bubble now. If the SegWit2X vs Core issue is resolved in a good way, we may see the price double.
Apropos Moores law, I have an anecdote ... I was doing some DBA work in my previous work place. We had one central database server, and we were always pushing it hard. The database server was constantly "bleeding edge" - the very best "ordinary" server we could get. We often we had performance problems - and most of the time we solved the performance problems by upgrading the hardware. We were efficiently surfing on Moores law to keep up with business growth and growth of database size.