They admit to a very small drag.
They admit in the small print that
"technically there is some resistance"
and I think the word "technically" is redundant.
If there is some, they cannot say there is none.
In practice I suspect there is no noticeable resistance because it takes out
so little power.
I prefer the national grid providing the energy for my batteries for my
lights, rather than me.
But the flipping-magnet scheme is very clever
It doesn't look a very efficient way of converting mechanical energy.
Their animation looks very vague indeed.
Perhaps deliberately so.
and as they point out there's very little mechanical
power required, and no friction like wheel-contact generators have.
I like it.
I got the impression that LEDs do appear very bright light sources when you
look at them directly, but they don't seem as good at illuminating the path
ahead due to their lower light output. I think the human eye can't perceive
a great difference between two point sources near the upper end of perceived
brightness very well.
I heard it is also a myth that LEDs are more efficient than filaments.
Efficiency of the latter varies with voltage and current.
Low voltage bulbs (e.g. 2.5V) have to pass the most current, so the filament
has to be thicker, hence there is more heat conducted along and out of the
wire. And also more resistive losses from wiring and internal resistance of
the batteries. On the other hand they can't make the filament too thin
otherwise it will break too easily. One can buy bike light bulbs designed
for 6V batteries, but I have not seen any for 12V.
Filament efficiency also drops as the battery voltage drops below the bulb's
rated voltage: the filament does not get hot enough to radiate visible light
very well, but still manages to radiate energy in infra red - depleting
batteries for less light output. Nickel cell voltage droops worse than
alkalines or zinc-carbon.
