Except for nodes extremely close to a HFC hub or head end, the overwhelming majority of the distance traversed through a HFC plant is over literal, optical fiber strands, and with remote PHY equipment the footprint of glass is only expanding. Also, for RF clarity reasons, the termination of the fiber in the street is almost guaranteed to be under a mile from the modem in the premises.
Extending this to claim connections feeding an access network implies the same logic indicates a bad-faith argument.
I have no interest in the legal or marketing aspects of this discussion, but from a purely technical standpoint, it’s entirely accurate to claim that HFC truly is fiber powered, and given the analogue lasers involved, there’s a hell of a lot more power fed into those fiber strands.