tl;dr - Average utilitarianism seems to have weird implications if we're averaging over time, instead of just over people. Is this discussed anywhere?
If we consider whether we'd prefer a society of 1 million blissfully happy people versus 2 million merely very happy people, we're in the realm of typical population ethics. However, if we instead ask whether we'd like to have 10 generations of blissfully happy people, or 100 generations of merely very happy people, it seems like a different question - not because of discounting, but because we might want to aggregate over time even if we average over people alive at any given time, or might want to average over time even if we sum over people alive at any given time, since these seem like conceptually distinct questions.
I suspect there is an interesting set of questions here, and wanted to know if it has been discussed, or if there are specific reasons to differentiate between these cases, or not to do so. (There are obvious implications for the long term future, but I'd prefer to ask about the population ethics for now.)