I've been wondering if SpaceX is going to develop a Starship "3rd stage" system for direct to GEO launches without refueling. Moving the Starship all the way to GEO and back just to get a satellite up there seems pretty wasteful compared to a relatively small expendable "stage" to move from LEO to GEO.
For better or worse, it is very unlikely a satellite destined to GEO is ever going to launch into anything other than GTO. So, whether a two stage solution gets you there or you need three stages (which, given SpaceX's philosophy, I doubt), it's still heading directly to 23k miles. More likely for BFR to GEO is a second stage that's sort of analogous to the Tesla skateboard: For non-human launches you put a 'regular' deployable fairing on top, for human launches you put the spaceship on top and call it a starship.
Interestingly, the above GEO equation changes if you start talking about in-space manufacturing. It is most probable that a manufacturing facility be placed in LEO (you don't want to pay to send your entire factory to GEO or wherever), so at some point you will need to raise your finished product from LEO to wherever its going. That's the point when you can start trading expendable or reusable tugs against on-board propulsion. For GEO the latter still probably wins because you need a propulsion system either way and so you might as well have the on-board system do everything, but there's probably some hypotheticals that have a tug win the trade.
If it were cheap enough, communications satellite companies might be customers since it would get them on station faster so they could earn more money.
I know I’ve broken recorded this concept already, but we're unfortunately far from a practical solution that's 'cheap enough' and, just from a basic physics perspective, 'faster' is a very unlikely case.
The only place where 'faster' comes into play is when a spacecraft's propulsion system is low thrust (so, electric) and thus requires a months-long circularization from +/-GTO to geosynchronous. BUT...the main reason spacecraft have electric propulsion in the first place is to allocate as much mass as possible to revenue generating payload. You don't want EP because it's more complicated and [at least right now] more expensive than chemical, but when you're mass limited the extra expense is offset by the additional revenue. So...when you buy a BFR and your paranoia over mass goes way down the decision to go with cheaper chemical propulsion is pretty easy, and so the ability to get on station "fast" is all but a freebie.