The excess power was 5% of the input power, and (as shown) roughly proportional to input power.
McKubre showed many other clear examples with lower input power. In some cases, output was 3 times higher than input. In other case there was no input (in heat after death).
However, this ratio makes no difference. Even 5% has a high s/n ratio with the equipment used in these experiments.
I selected this particular graph because it was convenient. It happens to be on my web page. I suppose you realize McKubre published many other graphs with much larger output to input ratios. If you do realize this, why did mention this non-issue? Are you trying to give readers here the impression that all of McKubre's results are at a low ratio?
That makes artifacts possible, and arguing that none such could exist complex. Hence not clear cut.
No, it does not make artifacts possible. Artifacts are ruled out by using good calorimetry, calibration and so on. The fact that output rises in response to higher input is not an artifact at all; it is a known property of the effect, and the cause of it -- higher temperatures and higher current density -- is well established. It is predicted by McKubre's equation.
Hence, it is quite clear cut. Your thinking, on the other hand, is muddy. You are ginning up non-existant problems out of thin air, and you are ignoring counter-examples such as tests with much higher output to input ratios. You are calling these figments of your own imagination "problems" and "artifacts." They are no such thing.