So there is a feedback loop here - if prices go up, it benefits the hashrate and thus the network security (and quality) which in return provides support to justify the prices.

This relationship is good on the upside but if the effect is the same on the way down, in such a cycle it would be viscious. Especially so if there was persistent price drop.

But just as you pointed out, 2017 peak did not cause such a hashrate problem. While as per your article it seems that backorders prevented that, I do not find explanation satisfactory for the almost 2 year downcycle.

Perhaps even the USD 5000 mark was sufficient to mine and the the USD 20000 to USD 5000 fall was not that important.

Yet the drop beyond USD 5000 had an impact on the hashrate that is observable.

Thank you for the article..