This is an excellent point. I'm curious to see how that pans out. My guess is that since Bitcoin transaction fees are supposed to be set by the market, they'll stay competitive. If mining is unprofitable at the rates credit cards are offering, fewer people will mine, which makes mining cost less energy, leaving more profit. Eventually an equilibrium will be reached.

The mining network shedding capacity is a bad thing for Bitcoin.

If 90% of the network is powered down to reduce transaction fees, that gives someone the opportunity to buy up 20% of it on the cheap and corrupt the network. Bitcoin has to use as many resources as possible to stay secure.

This is the biggest practical flaw I've understood so far, but it still seems fixable. If Bitcoin becomes uncompetitive with traditional payment systems and mining activity decreases, miners could collectively decide to alter the mining reward.

The problem with keeping mining rewards high is that Bitcoin is supposed to be deflationary.

Right now the mining rewards being minted increase the Bitcoin supply at a rate of 11% per year. If demand for Bitcoin were constant, this would amount to 11% inflation. Demand for Bitcoin is currently increasing, so it is net deflationary, but this only lasts as long as the demand for Bitcoin is growing constantly. If you kept mining rewards high after the demand for Bitcoins had stabilized, you'd have five times the inflation of the dreaded US dollar, as you paid for your infrastructure costs through inflationary taxation.

(Which Benjamin Franklin and I actually like, but that's another issue entirely.)