The last essay in the G. A. Cohen collection shows up an age-old issue in moral philosophy, namely, the recurrent belief that any answer to the question 'Why be moral?' needs a metaphysical authority to back it up. This might be thought of as a leftover from religion, but in fact extends far beyond religious teachings.
In a sense Cohen himself takes that stand when he argues - in criticising Christine Korsgaard - that if we drop the notion of morality being required by reason as such as Kant would have it (or by God), then we cannot treat morality as like a law that we must obey. Instead, it has to depend in the last analysis upon our personal identity, such as feeling bad about ourselves if we behave immorally. The problem is that for many people this raises an alarm about anything goes - Cohen points out that we cannot assume that everybody else will feel badly about doing something that I would feel bad about.
Maybe the way around this sort of problem is abandon trying to derive our moral values in the way we might derive a logical or mathematical theorem, but consider practical 'reason' in terms of taking notice of the concerns of other people as well as ourselves. The result would not be certain, it would be loose in places, but it would have a source outside our own personal whims or idiosyncrasies.
In my past I argued that a loose source of morals becomes tighter with the growing destructive power of technology, making our failures and transgressions more dangerous. I have doubts about that now, although global communications and other technology (not least weapons) give at least some moral appeals, such as climate change, a force they would not have previously had. Again, no certainty here but perhaps a basis for rough and ready rules that people would ignore at peril.