So the case of friendship is helpful as a jumping off point for understanding non-consequential moral reasoning. But I want to note that non-consequential reasoning is quite ubiquitous, and almost everyone has strong non-consequentialist intuitions. For example, take another well-known thought experiment. You go into the doctor's office for a minor illness, and while you're in the waiting room, they discover that there are five other patients on the brink of death, all of whom could be saved by a transplant of one of your organs. Should the doctors surreptitiously kill you, and then harvest your organs for the other five people? Clearly, the world in which at the end of the day five people walk out of the hospital alive, rather than just one, is in one sense a better outcome. But almost everyone agrees that this would be abhorrent.
Of course, consequentialists always have ways to try to explain away these intuitions. It's often said that a policy of doing such a thing would create such harmful effects, in terms of people not trusting the doctors, and general societal unease, that the consequences actually count against such action. There's something obviously true about this. But of course we can always alter the case, such that it's unlikely that such a circumstance will occur again, no one will every find out, etc., and I think we still retain our initial judgment that it would be the wrong thing to do. And, at least for me, the initial reaction to the case is not, "The policy would have terrible consequences that outweigh any potential benefit," but rather that "The doctors have no right to use me in that way." Of course, a consequentialist would likely provide an error theory about why our intuitions are confused in this way. But it's hard to think of my reaction to this as a kind of confusion.
There's another response, which hopefully I'll be able to give a longer treatment of later, which is to distinguish between act-consequentialism and rule-consequentialism. Act-consequentialists assert that the consequences of the act are all that matter in our moral evaluation, whereas rule-consequentialists think we should act according to the rules that would have the best consequences. This is an enormous topic, many different versions of both theories have been put forward. Nevertheless, I think rule-consequentialism, at least in terms of a moral theory in which justification comes only from outcomes, is inherently flawed. There will always be slippage back towards a more act-consequentialist view, which falls prey to the type of counter-examples I'm discussing.
I find this case helpful because it also helps to disabuse one of the notion that these questions aren't important. Obviously, this exact case is rather implausible. But our reactions to it will help us make sense of how we react to other cases in the real world that might share similar features. And if it strikes you as important that you not be made into a forced organ donor, and that others not be made into forced organ donors as well, then ethical philosophy has some importance. It is good for us to think about and discuss the salient features of these cases so we can better understand what matters, and also what we have most reason to do.