Just the 1 LED?
Not 1, about 24X26 of them...., but only about 30% are lit at any time.
Assuming each LED is ~5mW LED, if all the LEDs were lit @ full power it would mean ~1+W of power.
About 30% of them are lit, at best, and most of the one's that are lit are dim (dim=10% of power), so I would say we're seeing here 50-200mW of electrical power going into the LEDs, at ~ 1 meter distance.
The spot seems like a ~100 cm^2 spot, assuming some 25% conversion efficiency, you get 2-8mW/cm^2 of sound, or ~135-140dB @ the receiver.
@paulReynolds can probably estimate this much better.
There are so many unknowns now in what they are doing, and inconsistent reported statements from them, it's getting hard, even for me, to make good estimates as to power levels, however I'll return to the point I made earlier that I had assumed that the LED detector panel was not 'self powered' and instead had batteries to amplify any incoming signal. That would be pretty simple to have a board with a M by N grid each with an amplifier, a Murata receiver on one side, and an LED on the other. If it's self powered you'd see that really quickly with distance. They also wouldn't be doing demos where they use a small panel to light a single LED, as I heard they did at CES in the private show, if each Murata could light a single LED.
I'm not even sure what they are doing now, it just doesn't make sense to me, not even in a "lipstick on a pig" way. If you can do 1 Watt at 1 meter, show it, it's 33x better than Energous. If it's certified safe, release the study and prove it. If it's legal in a country you can say "under <insert dB limit here> at all points". If you can't charge for shit (kinda important), if it's not safe, and if it isn't (or you're not sure) that it's legal, then you might want to rethink your business plan and pivot PDQ because things aren't going to get better from here.
Basically, IMO, if they had what they're apparently claiming, it makes complete sense to go public with it all. That they don't is pretty telling.