Sort of.
As noted a while back,
Kessler is the vulnerability with megaconstellations, and that vulnerability increases as the size of the constellation goes up (and as the altitude comes down). At the massive quantity of Starlink It doesn't take a lot of kinetic impacts to all but nullify the use of an orbital shell--and if the satellites move to a different altitude to avoid the debris clouds the aggressor just does a rinse-and-repeat attack.
The important factor here is that the aggressor doesn't actually need Kessler to take out a material number of satellites to have a significant impact on service. One of the major differentiators for Starlink is that the conops/COLAs are highly automated--that's what actually enables them to fly so many satellites in close proximity versus old school human-in-the-loop satellite operations. The downside to that implementation is that there are TONS of automated maneuvers happening every day to ensure the satellites don't run into each other (or other objects crossing their orbits).
When a satellite is performing one of these maneuvers it, at best, provides reduced service. This is partially a function of available power (electric propulsion requires hundreds of watts) and partially a function of pointing (the satellite has to orient the thrust vector for the COLA). That's no biggie when there are a gazillion satellites because there's plenty of other satellites to pick up the slack of the relatively few satellites that are COLAing at any one time. In the Kessler scenario though, one could easily imagine orders of magnitude more COLAs in the constellation, to the point where--assuming processing power and predictive algorithms can keep up--a material number of satellites are constantly dodging debris rather than providing service. So basically an aggressor just needs to keep the rate of attacks out ahead of the rate at which the constellation can maintain a useful number of in-service satellites.
The OTHER big thing with satellite service is that the user terminals (regardless if for a LEO constellation or a GEO sat) are pretty obvious beacons. They operate in a strictly defined frequency range and a predictable power level. An aggressor could quite easily and quickly pinpoint [active] terminals with a pretty standard airborne platform and then execute strikes on those terminals. Obviously that's an extreme scenario that we all hope doesn't come to fruition ever, let alone in this conflict, but imagine if Vlad authorized--and advertised--airstrikes specifically on Starlink terminals... He'd just need to take out a handful of them before most everyone else shut theirs down.
