[Plus, we both measured and estimated heat losses from the calorimeter chamber walls, as described on p. 8 of the ICCF21 paper.]
One of us isn't understanding the other. You did this on the 3kW output reactor?
Mizuno did this with all reactors, in all tests. So did I, since it just involves noodling with the data at long distance.
He did not "do this" during the open air tests in his house (Fig. 1). How could he? All of the tests I described above can only be done in a calorimeter.
I suggest you downplay the open air test in your mind. It proves little. It was only done because the calorimeter cannot measure a heat above ~1000 W. Other data from the calorimeter point to the open air test as being right, but this is a crude method that tells you nothing about the reaction. A test showing 250 W excess in a calorimeter is far more believable than a open air test. It is far better.
I find it a little odd that you were put off by Rossi's open air tests, yet you are so enthusiastic about Mizuno's. Rossi should have used a proper calorimeter. A large, air-flow type is probably the only kind he could have used with such large reactors. If he had done this, it probably would have revealed that his reactors did not produce any excess heat. That's the kind of confirmation you want. You can't get it with an open air test. Okay, you can, sort of, but it is much less definitive, and it may well be a mistake. It is doing things the hard way, for no reason.
Some cells and some cold fusion devices are, themselves, calorimeters. All hot water heaters are. They happen to have the right specifications for good calorimetry, because those same specifications are what you need to heat water economically and keep it hot. For example, they do not lose much heat from the walls of the reactor, so nearly all of the heat is captured in the flow. That's what you want in hot water heater and in a calorimeter. The on-demand electrically heated type of water heater makes an excellent flow calorimeter.