It's interesting how an anecdote like this:
"There is also noticable noise if I am doing physical activity and it needs to blow the CO2 out more quickly"
significantly increases my sense of the strength of the evidence base that a device like this is actually doing its intended job. The idea of a repeatable (and repeated) experiment based on principles that we know a priori must be correct (that our lungs exhale more CO2 than they inhale, and at a greater scale during vigorous exercise)
Somehow it's more convincing than if poster had even said, 'I set a CO2 generator up and looked at the ventilation device's activity indicator and saw it increase whenever the CO2 generator's output setting was increased".
Rather, one hears the device working harder, when doing an activity that's familiar to all of us, in a repeatable way.
Also, the poster didn't have to tell us that lungs exhale more CO2 than they inhale, rather leaving that to our internal models of how that works. If they described that overtly, it would somehow already make me slightly more skeptical of the overall assertion of the device's efficacy.
I think the statement is made more powerfully convincing precisely because it leaves it to us to infer the implications of the obvious bits. When it's our own subconscious reminding us of the priors (1. lungs expel more CO2 during vigorous exercise than normal, and 2. many machines are noisier when they have to work harder to accomplish the task they are designed for), we intuitively trust that our own model of the situation is accurately informing us more than if someone else had to tell us that those underlying assumptions were correct.
I wonder how other demonstrations of innovations of less-apparent efficacy can be directed towards presenting their evidence in this more intuitively convincing framework. (Note I'm not saying I have particular reason to accept the poster's description without further diligence, merely that it 'feels' more convincing in a powerfully intuitive way, which is what I'm after here).
Like 'dependent variable X should be something familiar to everyone, drawing on principles that most people would have already ascertained a priori, and, 'dependent variable Y should manifest as an un-instrumented artifact of the device in question, one that can be intuitively perceived with as few opportunities for technical obfuscation (which could otherwise give rise to either intentional misinterpretation or simply misapprehension) as possible.'
(in the example, this is the device making more noise).
Anyway, I found it very satisfying to read that sentence from the poster.