Hmmmm.....Remind me, how many Watts could optimistically be saved by the use of this method again?
All in all it reminds me of the much more simple method I used for years which is quite similar to the DIY method suggested. It does not include the use of lacquer thinner or denatured alcohol but uses the same jar shaking and a number of passes of citrus based degreaser, industrial degreaser (not sure what it's made of) and plain diesel oil (lamp oil).
The chain is then wiped dry by running it through a cotton cloth until there are no marks left on it. You'd still be surprised to see the amount of dark traces of grime on the cloth which is where I wish I had a US tank.
It is then put on the bike and lubed with Boeshield T9 (which contains parafin) and sprayed with dry PTFE for a few seconds whilst rotating the chain.
Boeshield T9 alone is actually adding friction. Adding PTFE makes it roll much more freely which is why I decided to abandon the parafin idea and use a lube charged with microscopic PTFE particles.
Not sure how good that method is but let's just say I'm rather skeptical about any method claiming to improve on it in the multiple Watt range.
Does a quick connect costs watts (even a fraction) vs. use of joining pin?
I suppose that would very much depend on its manufacturing tolerances, its state of lubrication and the mileage it has seen.
Maybe someone ought to build a chain using nothing but quicklinks and measure the resulting drag? Who says bicycle chains need rollers anyway?