The only way to man in the middle to have such a large percentage of hash power is to either a) attack a pool (which people would notice when their withdrawals go nowhere), b) attack a large number of nodes, which must have enough combined hash power to mine four blocks within three days for people to notice (I think it is unlikely for Bitcoin point of sale nodes to have significant hash power), or c) the attacker himself has 1% of the hash power and is diverting it to conduct a man in the middle attack against one single person (as opposed to a major retailer who has a round the clock IT staff). In order for a large number of nodes to be attacked, it must be by someone who either is a state actor or an ISP, at which point you've already lost.

It's really simple math, it require on even the most optimistic estimates a tenth of a percent of the total network hash power to mine 4 blocks within three days with good luck. Or maybe this single person is on vacation, then it would take a hundredth of a percent of the total hash power over two weeks. I think very few people even have a hundredth of a percent of the total hash power, which goes to show how secure the network is, and how little my proposal would weaken network security. I'll concede that difficulty could be reduced only by 80% if only four blocks were mined in 3 days, which would provide sufficient margin against these proposed man in the middle attacks, because block-chain growth would be noticeably reduced.

But I repeat myself. Repeatedly. I wish you would understand my points. I'm making a good faith effort to provide an original idea before it's possibly too late. But fine. I have nothing more to add, and it's the holidays.