Yes ok so...HOW? haha, that was my question! HOW can digital cables be different? It seems to violate the principle of digital data transfer. Unless there are missing bits and no parity built into the system?
It could be bandwidth, it could be reflections due to incorrect impedance matching from either the source or receiver side or the cable impedance (for longer lengths).
Because of this the 'signal fidelity' of the, in essence 'analog' nature' of the 'digital' signal can be below par which may or may not become an issue when reconstructing the digital info (jitter or bits toppling) because the receiver cannot make the correct 'heads an tails' from the incoming electrical signal.
The culprit here is not the cable as such but actually
the receiver side that is not as immune as it should be for degradation of the electrical signals.
Then there is common mode issues, power/ground wiring being too high in resistance etc.
So in some cases there could be issues with one cable but not with another cable. The common audiophile 'differences' in sound quality often seem to be the same as if it were an analog transmission which cannot possibly be the same due to the nature of the signal (both are electrical voltages changing over time)
Consider coax cables in analog TV where one cable could have less noise with certain channels than others between cables. It could be that the cable had a poor impedance or it was bent once. Also with digital TV the same thing can happen. Cable one is fine, cable 2 has 'blocking' or signal loss or stuttering picture or sound.
In these cases similar issues are present in the cables but work out differently.
Then there is the transmission mode (handshake or pushing) where packets are not resent and have to rely on redundancy/error correction or loose (parts) of packages.
It is incredible that it works so well in practice with a wide range of gear and cables anyway. That's the beauty of digital transmission.