I'm not convinced that you actually do have to make too many assumptions. The claim I'm making here is that bitcoin has had a steady and predictable increase in mining demand as the currency has grown, but that iff bitcoin becomes solely a low-volume high-value settlement platform (instead of a buy-coffee high-volume low-value blockchain), there will be a demand shock once the block rewards go to zero. You can reason about this while keeping the hashrates and cost of achieving that hashrate abstract.
If volume grows, this argument is moot because there will be plenty of transaction fees to gobble.