You wouldn't need an infrastructure company, nor would you need a big capital expenditure. The point is that the infrastructure would be state-owned and available on a non-discriminatory basis already. If demand exists, a competitor could bootstrap by buying fractional bandwidth from the last mile infrastructure, fractional bandwidth from a peering provider, and connecting the two.

The problem at the moment is that huge capital expenditure is required because I need the majority of a street to be interested before digging it up becomes economically viable. This problem would go away.