This is a very good point. At some point in the future the mining power will need to be paid majorly by the transaction fees. This depends a lot on how high the

This is a very good point. At some point in the future the mining power will need to be paid majorly by the transaction fees. This depends a lot on how high the price can rise, but lets assume gold market cap. Then it would be maybe 3-4 halvings = 12-15 years until the share of block rewards in miners compensation will decline and the transaction fees have to compensate for it. In my view this is the biggest long term security and usability issue because transactions will probably cost hundreds to thousands of current day dollars assuming steady or slightly rising power usage = tech-innovation adjusted hash-rate. Hundreds of dollars are too much for the average joe to regularly open and close lighting channels, so we might see some kind of centralisation. The big question is if bitcoin can provide enough value as a international settlement backbone to fund the biggest pool of sha-mining hardware in the world. If an alternstive usecase for sha-miners comes up (be it blockchain related or not) that would also be a bigg security blow (but I don‘t see this on the horizon for now)