Because as I understand it L5 is parity with human capabilities. Until you clear that you're underperforming in some measurable way with what even mediocre human operators are capable of.
If a robot taxi can drive way better than a mediocre (read: every human driver when drunk/tired/angry) human can, but is geofence limited to a single area, that is an L4 system.
As a fellow driver, I don't care that they can't drive out of that area, I care if tf they're a safer driver.
It would absolutely be the wrong bar if there wasn't a steady drum beat of calls for ubiquity. "I can't wait until the majority of vehicles..." etc etc. This kind of uncritical optimism has already lead to some pretty wild outcomes with half-baked automation.
It gets blurry towards the edges. There are certainly roads I should not be driving on that competent 4WD folks are (in the right vehicles). I;ve also driven in winter conditions that felt very sketchy to me that I may have felt the need to drive in for work or other reasons. You don't always have the option to just Nope. I'm going to stay at home.
And, of course, if you otherwise basically never drive, you're probably a hazard in those marginal conditions.
They systematically get in less accidents, the accidents they do get in are not their fault, and they tend to be less dangerous when they occur. As I understand it, that unambiguously means outperforming.
We don’t need to remove every manual car in the world overnight, but this seems like a clear win, at least in the regions they operate. They don’t need to cure cancer and drive in every road condition at all times in all places and also not ever make any mistakes at all to be an improvement. “L5” is just a label, which I guess doesn’t apply, while “safer”, “less accident prone” are different labels, which do apply.