It really depends on the vector of attack, and this is why it's important to understand that space systems are made up of those four segments.
If it is a kinetic, physical attack on one satellite, then our concern is what the debris is going to do to other satellites in space. In fact, as I mentioned, the much greater threats are ground-based and link-based like jamming a communications satellite link; spoofing and sending a false navigational link so you don't know where you are, where your adversaries are or where your target is; dazzling an earth observation satellite so it can't gather the information it's trying to gather; and cyber-attacks. These non-physical, non-kinetic attacks are much more prevalent, much more useful, in fact, and much harder to attribute, so they're also much more effective in times of tension and warfare.
In those senses, you're often attacking a different part of the system than just a satellite. The reason people keep talking about distributed architecture and what Spacelink has done is that, when you have many satellites, it no longer matters if you target one single satellite or one single link, because the other satellites can take the place of that, and service can continue to be provided. The more complex you make your architecture, the harder it is to attack, which is a defensive mode.
I don't know if that answers your question.