Keen is mistaken about people not spending Bitcoin, maybe they are "idiots" but why is an economist wasting his time making psychological claims? Why does he bother thinking others will do what he thinks is prudent? There is plenty of evidence that most people have high time preference (they want something now, rather than waiting for it), the amount of debt and growth of debt in the economy is a perfect indicator that people would rather get something now and pay more over the long run, rather than save up money and buy something later at a discounted price. Obviously this is slightly distorted by monetary policy which lowers interest rates and can artificially drive up demand for loans, but that also doesn't account for the fact that things keep getting cheaper, technology gets cheaper, all we do know is what we see in the Market and that people prefer having things now rather than later.
Also BitPay has seen sales growth of 328% since 2016!
https://blog.bitpay.com/bitpay-growth-2017/
On the other hand, I think he is right that the price has to come down (or at least not rise as much as the total market capitalization of all cryptocurrencies) once people realize transaction fees are higher than competing cryptocurrencies. This is more to do with public discover of these alternatives and not so much about preferences. Which to me is good if it happens, that means Bitcoin will have broken any inhibition they had and people will start acquiring and using other cryptocurrencies.
I mean we can kind of see that now, even though Bitcoin has risen a lot this year it has actually lost relative to other cryptocurrencies going from 85% to as low as 40% of market dominance before coming back to 56%:
https://coinmarketcap.com/charts/
What that means is Bitcoin at $11,117 would have been $16,874 if it could have kept it's dominance at 85%. So the Bitcoin bubble popping might just show up as slow growth relative to the rest of the market, not necessarily a crash in the price.