First, was your battery sufficiently charged that it would always want more current (e.g. controller in bulk charging) during all your tests ?
But, yea, for short wiring runs the difference may not be much, if anything - if you look at the charts some manufacturers provide, the higher the voltage input, the less efficient the controllers are. So unless the wiring loss exceeds that loss of efficiency, things might even be worse at higher voltages.
The whole low light tale doesn't make any sense, and I've never heard any reasonable scientific explanation for the belief. But you're right - at such low levels the available power is mice nuts.
When panels were $5+/W, maybe it made sense to spend money to eek out every coulomb. But at current prices, it makes more sense to just add another panel, space permitting. But going a step higher higher in voltage to get back 3% wiring loss but lose 2% in controller efficiency - meh. If that 1% makes a difference, you're in trouble either way.
I think the best reason for higher voltage MPPT is that the large higher voltage/wattage residential panels are much cheaper per watt, if you can get them locally and not have to pay shipping.
