Sorry to do some spouting...
Interesting Boro... I did get my Moli from an eBay source and later found out they were very old, so maybe my one cell is a dud. From Electrolumens, I would trust more - definitely, so maybe they are very good. I'm still clueless what an X3vn is though. Also interesting you got them to fit in X3. I assume an "X3" is a pretty high end light, so maybe it was designed with large high qual springs to accommodate 700's.
Ohhh - almost forgot this issue... It's deceptive at times, depending on the design, but long cells compress springs more (obvious), and what happens is the resistance of the springs is reduced because it reduces the travel significantly vs uncompressed. So it very deceptive, but practical (!!), that long cells will perform better than short cells sometimes. You can easily verify this (I've seen it many timed), when you do a tailcap measurement in a typical direct drive single cell light, the harder you press down on the cell, which compress's the spring, the higher amp readings you get. Higher amps of course is more output in direct drive.
Since vinh has made it clear in the past he did not believe in copper braiding springs, then the only alternative is using springs highly compressed. Unfortunately there doesn't seem to exist a cost effective spring that has high conductive properties, at least high enough to radically reduce resistance when uncompressed. Using copper or silver even, makes the spring too soft for practical use. Berrylium copper is better, but very rare to even find one in the size we need, and even still, not as low in resistance as straight copper or silver. Most springs are steel with a conductive coating, could be gold, silver, etc., but it's a very thin coating and has lots of voltage loss's at the amps we are running at, considering the distance the electricity has to travel (uncoiled spring).
Again, maybe practical in real world use, a longer cell is the answer to get higher output, but not because the cell is better. This again varies greatly from light design to light design.
So to sort of conclude, vinh is testing in his specific modded light and came up with results very specific to his light. I think maybe he didn't use copper braiding or wiring over the springs, therefore the results will slant in favor of long cells because amps in direct drive will vary greatly on resistance, and compressed springs have much less resistance than uncompressed springs. Us budget modders, know about this and copper braid every spring we come across
. You can make an argument for high quality springs, but I've tried some, and they do not perform as well as copper braiding, so they will be better, but not as good as direct copper braiding/wiring.