Supply and demand, and elasticity in each, for broadband services vary dramatically around the world, in regions as small as a few miles.
He'd be crazy to stick to the flat pricing model he is proposing - it would be leaving so much value on the table.
As bandwidth gets oversubscribed, the 'fixed price plan' becomes the basic service tier, and he can charge extra for 'speed boosts' that make it faster. This then becomes a self balancing market - the more users there are in a region, the more will have to pay for boosts to get acceptable speeds, which effectively increases the price and makes some users consider other competing services.
Each 'speed boost' could cost $100/month, and be sold as 'it's just like having two starlink accounts and having the speed of both combined'. The dish could be color coded (white, bronze, silver, gold) to indicate 0, 1, 2 or 3 speed boosts as a status symbol.
There are places in poorer nations where $100/month is more than anybody can afford, and adoption will end up very low... Yet there are places in the USA where $100/month is a fairly low price and the network won't be able to cope with the demand - you would get more people online if you used carefully priced plans to create many different service levels (speeds, 'nighttime boost', etc) to optimally fill the infrastructure.
No economist would tell you to go for a flat pricing structure, no matter what your goal.
I think the only benefit is for ease of marketing/PR. You get more word of mouth sales if everyone knows both the price and service is fixed. Like the 99 cent hamburger.
At that point, the cost of a microwave backhaul, even if you end up needing a few repeaters, is a small percentage of the project cost.
That means even if starlink works out cheaper than the microwave backhaul, you won't see a sudden growth of these kinds of rural transmitters, since the economics for the phone company don't change much.