Each argument is correct. However if you want to trial e-cigarettes, it's very much like trialling desserts. One you like, I'll hate, and vice-versa. A trial of one
dessert would conclude that very few people like them.
What won't be mentioned by anyone using the data, a lot further on down the road, is that very few subjects would actually prefer tiramisu with custard, which is what you gave them. Some prefer ice cream, some apple pie, some want it with cream, some with custard. Some want it hot, some want it cold. A trial of cold tiramisu with hot custard is not a trial of desserts - 90% may not like it, and go back to candy bars.
E-cigarettes offer a huge range of options, with different types, and different models within those types; different types of refill (PG, VG, etc); different flavors; and different nicotine strengths.
That's why they work. There is a variety of options, something to suit everyone. A single option suits 5% of people or whatever. The rest would go back to smoking.
If the trial is locked to one model, then you have to accept that a good percentage will fail to stay with it as the model is not suitable for them. This has to be fully explained in the conclusions: "We would expect to achieve an additional 40% success rate if we offered a full range of equipment". Any other conclusion is a misrepresentation. In effect you are trialling cheese and biscuits, if that's what you offer - not desserts. You can't answer the question, "Do people like desserts?" if you only offer them cheese and biscuits.
And at the very least, you have to offer a full range of flavors and a full range of nicotine strengths - or it won't work at all. We already know what percentage succeed when given one option of hardware: 31%. Give them a proper range, and we get 80%, anecdotally. The difference is so large that it renders a restricted test virtually meaningless. You will get results of the same type as Eissenberg 1 - i.e. laughable and meaningless.
If you want to trial e-cigarettes then that is what you should do. Don't give cigarette smokers a cigar, or vice-versa. The results won't be worth the paper they are written on.
The problem is, researchers seem to think there is some sort of parallel between
tobacco cigarettes and electronic cigarettes. There isn't - every single aspect is different. The funniest results of all are obtained by researchers testing the vapor: every photo of their lab set-ups show them using the e-cig upside down, as that was how they tested cigarettes. E-cigs don't work like that. A cigarette might, but an e-cigarette is not a tobacco cigarette. An e-cig works about as well upside down as your electric kettle would, as it's a gravity-fed liquid-feed device. Thus the 'vapor' result is shown to include smoke from melting plastics...
Researchers have a big problem - they need to fit their methodology to the subject under test - not the other way round. Please try to learn from previous mistakes, there is no point in repeating them.
You could certainly run a meaningful trial using 4 different models, as these would suit most people. The question is: over what timeframe? Over 1 month and you'll be laughing. Over a year, and with one model only, you'll get about a 70% failure rate. We already know this. Look, here is the vaper's progress:
Day 1: 95% must have something that looks just like a cigarette and preferably tastes like one as well. They will give up within a week if given an eGo or Riva or something even bigger. They just want a cigarette replacement - and nothing else will do. Preferably an auto model.
Day 30: At least half have got sick of the short battery life and pathetic vapor offered by their mini, and want something better. They get a large-format like an eGo or Riva. They won't accept a large mod or a boxmod. Some will accept a manual model now.
Day 90: almost all have moved up to a large-format or mod. Manual is pretty much standard, they've realized that auto models are for noobs. Almost everyone has two different models, for different times of the day and so on.
Day 180: 75% are using a large-format or a mod. They are very, very picky about what they will or won't use. Just like you won't wear your wife's pink shirt and she won't wear your oily jeans you fix the car with. Hardly surprising - chacun à son goût. Everyone has several different models by now (I've never heard of anyone @6m with one e-cig - it just doesn't exist).
Day 360: Most now use a large-format or a big tube mod, or a boxmod, and with a
tank or a carto or a dripper atty (there is no kind of conformity). In other words, most would have dropped out if only a mini was available. And over 50% would also have dropped out if only a large-format was available, as they preferred a mod or a mini.
Some use 0 nic, some 18mg, some 36mg, a few even use 48mg or 54mg, which has to be DIY'ed as you can't buy it easily. Some use DIY WTA liquid as it's hard to buy. There is no conformity.
Some use 5% mild flavors, some use 15% average flavor, some use 25% of chilli extract that would rip your throat out. There is no conformity.
Oh, and people have a selection of models to choose from now. Some have 50 or more, all have at least three, and the latter are modern-day Scrooges.
A trial of e-cigarettes needs to be just that. In contrast, a trial of an eGo or a Riva (both are 'large-format' e-cigs) is just that: a trial of an eGo. It certainly isn't a trial of e-cigarettes, and it would be a misrepresentation to call it such. I think we have all had enough of misrepresentation.
Please don't get sucked into the junk science department, e-cigarettes are a major trap for the unwary and many have fallen in. You cannot later on say you didn't realise, after all. Tom E. had to start again from the beginning, to the sound of riotous laughter - but you have had advance warning.