One thing to remember is that ALL USB connections are powered.
That's correct and quite fundamental. The USB protocol uses the power on/off to do things like USB device resets and to start up the enumeration process, whereby the device introduces itself to the host and they negotiate what the device wants and what the host and/or hub says it can have.
USB ports have rules about what power can be taken from them and a unpowered hub can supply only what is allowed by the port into which it is plugged. Powered hubs have local autonomy on the power they can supply.
My best guess is that the issue is not the power side but the data stream side.
It could be both ... if the device says it wants 1A and the host or hub says it can't supply that, then the device will not operate or will operate in a lower-power mode.
It may be that external hubs (and I will call them that because as I said, ALL USB ports are powered from somewhere) with several devices connected were trying to pass too much data through a USB 2.0 port.
The device/hub/host will all also negotiate a throughput, but it will only be as fast as the whole chain can accept. I remember a time when thing like memory sticks would get an on-screen prompt that says they "may run faster if plugged into a USB ?.? port". Possibly that's now seen as "too hard for the average user", or maybe "Oh, they'll try another port anyway".
Shall we get into the conversation as to whether even passive hubs (with no wall power supply) are in a real sense "powered" hubs because the plug into the computer and thus are bus powered?
No, please don't. Better would be self-powered hubs versus host-powered hubs.