I'm not an expert. The evidence that USB2 was able to translate & carry usb1 is a good counter example to my premise, that the work was done & it worked out fine, and usb1 carries fine over usb2. For a while, I do not believe the USB transaction translator[1] was not totally commonplace: I remember articles specifically highlighting which usb2 hubs had this feature. I dont remember this feature ever causing compatibility problems, but USB1 had such a small feature profile, such limited concerns that it feels like this task was more managable. Once we start adding isosynchronous profiles & perhaps various other usb2 gimmicks, I imagine the transaction translation problem becoming much more thorny; I imagine trying to insure in fact that usb1, usb2, usb3 devices would all translate successfully would be much harder. I'm sympathetic in extreme to the notion that there should be a spec, that we should have tried to define something (and perhaps the specification without mandate would have been great, as usb2 did). But I also still feel like it would/could have been a risk, could have become thorny. That doubt doesn't convince me, but I want to acknowledge that doubt.
I also think there is some modest technical advantages. Making a hub that passes through usb4, but which also exposes a variety of simple peripherals is much easier without mandating the transaction translator. It's super convenient to make a hub that passes through usb3/4 but which offers a variety of audio, keyboard, mouse, et-cetera connecitivyt options, at a reduced speed: one can just drop down some cheap-as-dirt usb2 hubs in their solution & glue in those worlds of peripehrals. Demanding all connectivity happen over usb4 & usb4 alone would have mandates much much more expensive solutions. That said, I'm not convinced the savings are worth the cost. If we simply had mandated we do good, build better, I suspect over time we would have found the cost of doing better to be negligible.
[1] https://en.wikipedia.org/wiki/USB_hub#Transaction_translator