Carson speed is multiplier of best L/D speed, the latter is pure geometric thingy where lift/drag angle remain the same irrespective of altitude when the power is off, in nil winds, for a giving altitude you will find that best L/D speed delivers max f(x) = max(distance) while Carson speed delivers max f(x) = max(distance*speed)
When flying with power straight and level, you are using power to “recover” altitude such it remains the sams but you still get those optimal f(x) 
in nil winds assuming as long as your engine have power to fly those said speeds while keeping altitude constant, you will find that changing altitude does not change you MPG at best L/D and does not change MPG*TAS at Carson speed
In reality, winds changes laterally & vertically, turbulent air, TAS/IAS does not follow ISA and piston engines are not designed for those speeds on hot days, plus it’s too slow !
As proxy Vld is near Vref, Carson is 1.3*Vld, so you are looking for 2*VS and maybe +15kts/-5kts depending on winds
Carson speed is multiplier of best L/D speed
That multiplier is said to be a constant.
The rest of the above I don’t understand. I think some words are missing, and/or puctuation?
Therein it is concluded that MPG increases over altitude
As the air is thinner, pumping losses reduce because the throttle can be more open. But there are many other factors e.g. prop efficiency varies (and in a complex way). Engine efficiency varies too, and this is a prime reason why best MPG is not obtained at best-l/d speed (at such low power settings, mechanical losses are proportionally bigger, as can be seen from the non-equal 0 intercepts on the axes here).
Pumping losses were known to be very important way back in ww2 – see here and in the book on WW2 engine development mentioned here.
The rest of the above I don’t understand. I think some words are missing, and/or puctuation?
I hope someone else understands it, it’s highly technical and not for everybody…
There is the article Dan published, I suggest that as start
