The only way to test the hypothesis to see if smoking schizophrenics did better in illness and/or required less medication than nonsmoking schizophrenics. To my knowledge, no such survey has been done.
I agree with you that it would be very nice to test this hypothesis (if it can even be dignified as such) as rigorously as possible. Truth is very hard to come by, and much of what seems intuitively obvious to us, even after careful observation and deep contemplation, can turn out to be very wrong when subjected to deeper scrutiny/better ways of understanding phenomena.
That said, I do not think that renders speculation along these lines useless (as long as it is understood that it is speculative,) for a couple of reasons:
1) (And this goes to why I mentioned schizophrenics, and really to my questions about how we define "health") I do not think that your question ([do] "smoking schizophrenics [do] better in illness and/or requir[e] less medication?") is the only relevant question. You don't define "doing better in illness," but I'd imagine many metrics you could choose here might miss aspects of what I would call "the pleasantness (or lack thereof) of subjective experience."
Some schizophrenics, absent some breakthrough, will never really be able to function well in society, and will always require some degree of assistance (if not occasional restraint.) It is tempting to measure "better in illness" solely by these metrics, as they are to some degree quantifiable. It is my opinion that it would be a mistake to use such as your sole metrics. Even for people whose outcomes, in that sense, are not improved by smoking I think we should ask how much they suffer, in the normal sense of suffering, as a result of their condition. This is difficult to measure (more on that later,) but I think it not inconsequential.
Changes in the amount of medication required are also a bit difficult to quantify, and not necessarily relevant, as it is possible that the optimal amount of other medications would not change, even in the face of improvements orthogonal to the effects of those medications. And of course it is _very_ difficult to know if the amount prescribed is the optimal amount, and there are huge ongoing arguments over this.
2) I think we are unlikely to see a lot of studies seriously exploring the potential health benefits of nicotine, and even more unlikely to see studies exploring benefits that are "purely psychic" benefits. I think the latter term exposes something of a false dichotomy, but I imagine you know what I mean.
3) It's pretty hard to find schizophrenics who don't smoke.. hard enough that I'd suspect a confounding factor.
In the face of these difficulties I will continue to speculate, and to speculatively draw tentative conclusions I know are both tentative and speculative. I will not allow my understanding of the world to be limited to what has been peer-reviewed, but I do hope that I will retain a sense of perspective that keeps me from drawing firm conlusions where I should not.
That said, when I see people self-medicating I assume that they derive some benefit from doing so. They might not derive a _net_ benefit over time, especially when the benefit is short-term, and the cost is long-term, but there must be some benefit, right?
That such a high percentage of schizophrenics smoke suggests to me that the benefit is particularly large for them (though I can think of alternative explanations.) I'm inclined to think (and this is indeed speculative) that the benefit they experience is not entirely different from the benefit that led the rest of us to smoke in the first place. And I am inclined to think that that benefit is not a benefit to be dismissed lightly.
What are we aiming at, when we aim at "improved health?" Greater longevity is one possible benefit, and I'll leave that one alone for the moment. But, that put aside, we aim at a better subjective experience of the world, don't we? We aim at feeling better. There are plenty of treatable conditions with an easy to find physical basis that will make me feel like hell, but won't shorten my life much. They are considered medical conditions.
So, what if using nicotine makes me feel less like hell than not using it does? What if using nicotine eases some of the discomfort that being alive necessarily entails, without dulling the enjoyment it entails? What if the reasons for this are related to the reasons that almost all schizophrenics smoke?
In other words, I'd not be surprised to find that nicotine itself helped with some physically defined syndromes like hypertension and Alzheimers, but I'm also inclined to think that it might, separated from its traditionally disastrous delivery mechanisms, significantly improve almost everyone's subjective experience of wellness. And, absent increases in life-span and mobility, isn't that the end-goal of most of the practice of medicine?
I'd tell everyone to vape if it weren't for the potential for dependence. I'm willing to buy vaping being less addictive than smoking, but that doesn't mean it doesn't lead to dependence.