I always hear this statistic on how proper zipper merging increases traffic flow rate over no strategy at all, and I simply do not understand how it helps.

They keep pointing to how much of the upstream second lane is "wasted". But like, from a strict perspective of flow rate, is it really?

The bottleneck restricting flow is the reduced speed single lane. Put a vehicle counter on it. Assuming no one wastes time getting through whatever funnel point there is, this flow is consistent. The same number of cars passing at the same speed are getting through regardless of whether the zipper point was a few cars back or ten kilometers back. Unless I can hear an explanation on how zipper merging changes this I remain unconvinced.

Zipper merging still has unquestionable advantages that are obvious to glean, of course.

Putting the merge point as close to the blockage as possible minimizes the time spent in the shared lane. Flow is the same, but the overall time spent in the jam is averaged over all drivers.

That "wasted lane" does not, as far as I can tell, improve flow. But it does improve storage. If cars are piling up at the choke point, utilizing the full extra lane keeps the pilup from backing up as far down the road, reducing potential domino effects through the road system.

Zipper merging is fairer to all vehicles by promoting a FIFO processing order. No one in the closed lane gets screwed, everyone gets through in roughly the order they showed up.

It has lots of advantages, and is clearly the winner, but I fail to see how increased flow is one of them.

Of course, I'm making a lot of assumptions about perfect behavior of drivers, while this statistic is supposedly real-world empirical data. That suggests there are significant inefficiencies in real-world human driving, and that the zipper merge addresses them somehow. But I can't fathom what those are or why zipper merging is relevant to them.