if the resistors are missing, what does that do to USB communication?
Indeed, that is the question, and I >>suspect<< is why some notice
more reliable nanoVNA connections via USB-2 hubs than to USB-3 hosts.
Are they just for using the USB as a sink (load?)
USB-C connectors are mechanically symmetric.
In order for half the 24 connections to NOT be redundant, electrical "keying" is wanted.
1) I >>guess<< that a single resistor from one CC to ground is wanted in nanoVNA for electrical keying
2) I >>suppose<< the original nanoVNA design, with DP1 tied to DP2 and DN1 tied to DN2,
was intended to suffice as symmetric don't-care.
3) I >>suspect<< that floating CCs provoke some USB hosts to misinterpret nanoVNA capabilities
and attempt sending signals via DP1 & DN1 that differ from those on DP2 & DN2
or even try to employ some lines not supported by USB-2 and not used by nanoVNA.
As with FAR too many other areas of commercialized innovation,
USB-C promoters evidently prefer acronyms to clear explanations,
with several documents "describing" functions to be accomplished by a USB-C cable,
called EMCA (really??), which make sense (to me) only if implemented by devices.