Answer

In my opinion there are several errors in the analysis there. Take, for example, the case of preferring half a bar of chocolate now over a full bar in a week, while at the same time not preferring half a bar in a year over a full bar in a year and a week. Here, immediacy gives the person something of value. What is irrational about that? It is like buying a lottery ticket at a price for which the expected monetary return is negative (say, paying 20 shekels for it when the chance of winning a million shekels is 1 in 100,000). Is that irrational? It can definitely be rational, if the person is paying (and that is probably the case) for the excitement and the chance/hope of becoming a millionaire. Why is paying for that irrational, while paying for a circus is rational? A person pays for a mental state that he wants to be in—pleasure or hope. That is part of his utility function. Seeing this as an irrational step comes from identifying the utility function with expected profit. That identification is by no means necessary. A person who pays more is either foolish, or else, from his point of view, the gain is not only the expected monetary return but also the hope, and so on. Returning to the chocolate example: a person who wants to eat chocolate now gets immediate enjoyment, and for that he is willing to pay by giving up half a bar. Immediacy is part of the utility (the pleasure), and to him that is worth the loss in profit. But in a year there is no immediacy, because in any case he is waiting, and there he prefers the whole bar.
The same applies to texting while driving, not exercising, or eating too much. All of these involve utility considerations that are more complex than just health or life in the abstract. By that logic, it would not be worthwhile to drive even without texting, because it is dangerous. Yet a person still drives from time to time because he prefers a normal and convenient life over eliminating all risk to life.
In short, as in many cases, these analyses are very problematic because they assume incorrect things about people's considerations. In principle, it is very hard to show that people are not rational, because you can always argue that non-objective differences (psychological and not material-financial) also have value for them and enter into their utility function. True, one can always say that the person does not really prefer the immediate pleasure and is simply irrational, but it is hard to prove that. Even if you ask him explicitly what he prefers, and he says the money, or life, or the full chocolate bar, that still does not mean he is irrational. It may be that he prefers that only when the matter is placed right in front of him. Otherwise he represses it, and that itself is his pleasure. For that he pays or loses. But that is, of course, already taking things a bit further.
Rabbi Moshe Feinstein once wrote in a responsum that if a person acts consistently according to his own view, then he is sane (he does not have the status of a mentally incompetent person). Thus, if someone thinks he is Napoleon, and accordingly wears an appropriate hat and gives orders—he is rational and logical, even though his name is Moshe Zuchmir and he is not from Corsica. I am not sure I agree with him (more precisely—I do not agree), but his mode of thought assumes what I have written here: rationality pertains only to the form of inference. The starting assumptions of the inference are personal, and it is hard to judge them in terms of rational or irrational.