As he points out though, the efficiency of the units pales in comparison to a separate splitter and panels.
Hydrogen is a real PITA to work with because it is such a small atom it can pass through the spaces between atoms in molecules. So the materials used for things like o-rings and valve seats are almost never perfectly gas tight. Think about a pipe run and imagine hydrogen seeping through at every joint along that pipe assuming the pipe material itself can hold down not leaking over time.
To make it useful you need a LARGE volume at high pressures so basically you have to fill a water baloon with an eye dropper. The amount of energy required to squeeze that eye dropper hard and fast enough really cuts into the efficiency of the hydrogen generation. With seperate splitters that have their own compressors they're only seeing 50% efficiency and these panels weren't anywhere near that without compression.
It's an interesting idea, but it takes such a HUGE hit in efficiency to make a nice AIO package that I don't expect it to ever get much past a couple prototype installations.
Just my thoughts though, I didn't think dirt was going to be so popular when it first came out. ?