That is why Elon Musk says that success of Starlink and Starship is linked - there's no demand for it outside Starlink, but there's no demand for Starlink without Starship due to high maintenance costs. That's why he sent an angry email fuming about possible bankruptcy.
There should be enough demand for Starlink even with current Falcon9 launch costs.
Rationale below.
---
Remember that Falcon 9 is only a stepping stone to SpaceShip and SuperHeavy
If it works, SpaceX launch costs will drop significantly
Currently it looks like $28M for a reusable Falcon9 for 17tonnes to LEO. That's $1647/kg. But they charge customers a lot more than that.
With a reusable SuperHeavy/Starship with 100tonnes to LEO, they should easily be able to do $20M for a reusable launch. That's $200/kg.
---
For each Starlink satellite, these are the projected figures I can see.
REVENUE
Bandwidth: 20GBPs
Users supported: 400 (based on 50MBps per user)
Lifespan: 5 years
Current Charge per User: $99 per month
Total 5 year revenue: $2.4M
COST
Weight: 260kg
Satellite Cost: $250K
Launch Cost: $52K (@$200/kg with Starship)
Total Initial Cost: $0.3M
---
2 things come out of this
Theoretically those Starlink prices could drop by half and still be very profitable.
If you do the calculations, it is also still profitable with current Falcon9 launch costs of $1647/kg
Now, there are a whole bunch of assumptions in all this, but you can see Starlink services with current Falcon9 launch costs are very profitable.
---
And so far, Starlink has 500,000 users who have placed orders. That works out as $0.6Billion per year in revenue.
But there are many times this number of people who are willing to pay for a broadband satellite service for $99 per month.
NB. All these figures are publicly available.