alt Hacker News> A key difference is illustrated in TFA. To maximize fraction of people owning a home, you dislike people renting or living with a homeowner. A spouse that marries a homeowner, for example. It would be "better" if they broke up and owned their own separate homes. Or perhaps they could buy two homes, one for each. Or they could by two homes and rent a third to actually live in - same metric value.
Or they could... both own the home? Independent of the rest of your argument, this is a pretty weird hypothetical case to illustrate your point. Joint property ownership is way more common than any of the "alternatives" you're suggesting, so it's hard to take what you're saying seriously when you've dismissed the obvious way that this would work for the overwhelming majority of married homeowners.
Yes, we're not talking about reality, we're talking about the measures of it. A metric would fall down on corner cases, so identifying the corner cases is how you determine if a metric is a good one.
"% people owning homes" is maximized by joint ownership, as well, but the corner cases are interesting: It is lower with big families living together or when anyone rents.
On the other hand, the "% homes lived in by owner" metric falls down in a much more realistic way: When there are fewer homes and they are lived in by owners, you can still have 90% of your people renting apartments while this metric hits 100%