So it's pulsing at about 900 something hertz (or whatever they actually use) and transmitting between them. You're still using capacitors (where DC wouldn't, or would be even smaller just for voltage drop suppression), and as you even said you'd need slew control and other complex EMI solutions over normal power (especially at high power). It doesn't have to do with the frequency of the power either, it's the sharp edges that have to be suppressed (hence the above). Normal DC also does not experience inductive losses, which pulsed DC will (and greater than normal AC at lower frequency). Also look up the "skin effect" (related to inductive losses). Plus that's not including power factor, which may cause weird effects with your slew control!
http://suddendocs.samtec.com/notesandwhitepapers/simplified-pulse-current.pdfAs NiHaoMike said about transmitting between packets, this isn't very efficient for bandwidth. There are normal systems that use data over top of power, which have much higher bandwidth and can do all you described while using normal DC or AC power. Like I said, you'd never have to pulse the power (even to send data in between)!
So the only thing this has going for it is this so called "Safe Touch" fault detection. You can do short circuit and arc suppression on normal systems, it's not new. Explain exactly how it's supposed to detect a person across the lines in these off periods and turn off the power faster than you can feel it (at whatever Hz too)? Remember, a person is just a resistance, so across the lines it just looks like a load. Are they using impedance testing? And I won't be satisfied until SOMEONE puts their hands across it and PROVES it's safe. (at least with that saw thing they used a sausage, although for this a person is required because a sausage can't feel a shock)
More than 100W power over twisted-pair cables like Cat6. Cellular network providers are hungry for such. They state that big displays (TVs) are / can be powered as well. There are loads of applications that can benefit from cabling that is not regulated in same safety category as AC mains.
"They state that" I don't care what the marketing dept says, I care about science! Over 100W over CAT6?! First of all, is CAT6 even rated for 300V? I don't think codes would allow that, so you'd have to use mains cable anyway just due to the voltage involved! At 300V 100W the current is 1/3 Amp. CAT6 is (generously) 24AWG, which means a max rating about 0.6 Amp. However, it's pulsed, so due to having a duty cycle the instantaneous current is actually greater than 1/3 Amp! The cooling effects won't be much either at that frequency! It may not burn up, but it'll still get warm and the temp will have to be considered.
You can experiment with this at home too, just use a 100W audio amp and a slew controlled square wave at whatever their frequency is and feed it flat out into a speaker through a long piece of CAT6 for a long time. Now tell me the results. Did it get warm? What was the distortion?