Actually, Starm is onto something, but it’s less significant than a lot of other factors at play. He’s referring to Carnot’s Law, which states that the theoretical efficiency of a heat engine is dictated by the difference between the peak temp and the temp of the environment (or the heat source and the heat sink). The formula at play is (T_H – T_L)/T_H = maximum theoretical efficiency. T_H is your heat source or combustion temp, and T_L is your heat sink, ambient air temp, or wherever you have the heat flowing to. All the temps are in Kelvins (i.e. Absolute Zero is 0 and the scale is scaled from there).

Now, your combustion temp is going to remain somewhat constant once your vehicle is up to operating temp, and I assume that it will be the same in the summer and winter. But the ambient air at your intake and exhaust can vary from, say, -20F (or less) in Minnesota in the winter, to 100+F in many places. If your combustion temp is 1000F, and the environment is 100F, then you have 61.6% efficiency possible. That number only climbs to 69.9 if you go with -20F.

So ideally, you could improve your Carnot efficiency a tiny bit in colder weather, but a lot of other real world factors come into play, like increased oil viscocity, lower energy content of your gas, etc. that conspire to negate that effect.

As far as the cooler air in the intake by using the snorkel, and not actually cooler air due to weather, it could also give you a very TINY improvement in efficiency for the same reason (delta T), but it might be negated by increased resistance by the longer intake or something.