is it simply Vf * current consumed by the load for each of the four diodes?

Not precisely, but maybe I am nitpicking here. Power dissipated is the integral of the voltage across the diode and the current through it. If V and I were constant then yes, the integral is just V * I. But the voltage across the diode varies non-linearly with current, and in the case of a real load, ie with a filter capacitor, the current comes in spikes, and there is no really simple formula. You must come up with a formula for diode voltage vs current, then calculate ( based on load and on filter capacitor ) when conduction starts and stops and set the limits of the integral accordingly. Then do the integral, if possible.

But, having said that, Vf * I dc may be a good enough approximation.

The good news is that I dont think there is a real practical need to calculate diode power dissipation, just get ones rated for the correct inverse voltage and average current, and they should be OK. Others may comment, but I think that is a safe practical approach.

As for the heater center tap, as you say it comes into play with tubes with 12.6 (series) or 6.3 (parallel) connections. Or as acenter tap for connecting resistors or pots to ground for hum reduction.