> But if the business model would be to cater only for 10% hardest cases, you don't need the funnel, since people coming to you already have a $100 need, so to say.
Now consider that the people with a $100 need are going to depend on how good the free system is. If the free system sucks, it's a lot.
But if the free system sucks, lots of people are going to be complaining about it because they don't like paying $100, and then it might improve some. The number of people who need something else goes down from 15% to 8%. Then there are fewer complaints, but the paid system has that many fewer customers.
At some point there are few enough people complaining and the free system stops being responsive to their complaints, but the main thing that depends on is how much the government cares about making their system any good.
If the premise of embarrassing the public system when a private system can do better is successful, the public system could get arbitrarily good over time -- until the private system doesn't have enough revenue to do better.
You push that point further down the line by not forcing the private system to duplicate the public system's work.
> That's no different from entering any other established market right now. Yes, incumbents have an advantage, this is always so, always have been so and always will remain so.
It's not the incumbent's advantage which is the issue because that isn't the change. What you'd be doing is closing the door to new entrants regardless of whether the incumbent continues to exist.
Even if the market provided enough revenue for Intuit to continue if well-managed, if they were to become mismanaged or otherwise go out of business, or just do a crappy job while not actually ceasing to exist, no one could replace them because no one else could cover the initial entry costs while competing with the free system.