My understanding is that this is not exactly true. As I understand, they opened the heat exchanger side cover, and not the remaining unit underneath (The secret black box). There have been many comments elsewhere about its volume and contents. None is acutally known for sure. Yes it was weighed, with a bathroom scale, apparently calibration checked by having two blokes who "knew" their weight stand on it to see what it said. Now that said, a digital bathroom scale is not "inaccurate", but it is hardly a controlled calibrated device for science.
From what I understand, the black box is quite small. There were 3 of them. The biggest part is the heat exchanger. The non inspectable parts were to small to contain sufficient batteries or other means of providing the energy.
That does not mean that there are no other means of fraud, e.g via wireless energy transfer of some means (e.g via induction like the stoves do), or a hidden wire, etc.
On the bathroom scale- thing. You can consider that most peoples weight changes by up to one kg throughout the day (full blader, empty bladder, full colon, empty colon, etc). Then you can also loose a kg rather quickly, if you have changed your diet and had some exercise (you can gain it just as quickly too). This can go within a couple of days.
So you can assume 2kg tolerance. The device was 99 kg (?) so 2kg up and down would be roughly 2% difference.
The question is what is the signifficance of these 2%? Could one assume a hidden depot of 2 kg of hydrogen somewhere in the device maybe?
Would those 2kg have been enough to keep the device running in self sustaining mode for that long?
I have no idea. Maybe someone with more time could calculate that.
Make two holes in a tea pot, plug a pipe through the holes across the tea pot, fill the tea pot with water, bring to a boil under mild fire, inject as much water at room temperature as the wall of the pipe can stand.
Well not quite.
You have to turn off the stove after the water has been boiling, then inject the water.
We need a measurement of the quantity of heat transmitted to the secondary circuit, practically speaking the integral of dT times the mass flow rate. This was not measured with a reasonable accuracy, hence no proof. QED.
Maybe next time.
I might have missed it, but I thought they had a flowmeter attached ot the output, no?