That is incorrect. Evaporation is measured when calibrating with a resistance heater, when there is no electrolysis. If the amount of water that is evaporated did not equal the expected amount based on the temperature, that would be obvious.
Coming back to this. It is I think another sweeping generalisation. I did not assume this in Staker's case. I did the equilibrium partial pressure calculation - which is what you mean. It showed evaporation 10X higher than what he stated informally (private communication). The paper has no formal statement of what is expected evaporation - it is juts not mentioned. I also inferred (I think I am right) that the H2+O2+H20 vapour has to exit the cell in this case without condensation at a lower temperature and return to the vessel because the gas exit is at the top is via a capillary tube. There is however also in this design a splash-back for bubbles (as you might expect) inside the calorimeter boundary. So we hope and expect that no liquid can exit the cell (I guess an explicit check for that would be good since it would be indistinguishable from evaporation in terms of fill-up volume, but alter the heat balance). However - in this case I think Staker is right that you would need a lot of evaporation to alter the heat balance, and anyway this cannot make excess heat look worse than it is. So the lack of explicit checking here is Ok but dangerous because it makes all the experimental conditions less certain - and makes cross-checks for other things less easy.
Also, again in this case - the relatively low electrolysis speed means that the gases in the cell have time to reach thermal equilibrium - otherwise you would get less water vapour than expected from partial pressures. It took me a while but I actually did the check using web calculators for evaporation rate!
But I would not assume any of those things in a different experiment.
And, I would not assume an experimenter considered them all, without this being explicitly mentioned in the write-up.
Maybe all scientists are so good at this stuff that they always consider everything. For LENR (or any other extraordinary in replicated claims) no-one can assume that and explicit consideration of anything you rely on must be written up in the paper. And, generally, people, no matter how eminent, can make silly (or not silly) mistakes.
(I know you will say LENR is replicated - look at it from the POV of someone on the outside of the LENR community who needs to spend maybe 10 hours or so looking at a paper to see if they can identify loopholes - and if they find them will not be inclined to do the same for many other papers, and therefore will reckon LENR is unproven).
THH