I see we agree. On this small point, at any rate.

I wonder who may have an incentive to code up an alternative implementation? Maybe somebody who already has millions of dollars tied up in capital equipment - someone whose continued profitability requires making any optimization allowed by the protocol...