In other words, can this someday compete with (or even eliminate the need for) current broadband infrastructure?
Most likely not. The amount of spectrum required to replace current broadband users would be significant. The amount of coordination overhead for even managing just 50,000,000 satellite client transmissions would probably make the system unusable. I think the target here is less of a last mile replacement and more of a town-wide Internet uplink replacement for places in the middle of nowhere. i.e. the town would get one satellite uplink and then run last-mile connections to their own users.
In theory, satellite Internet could have any capacity you want by flying more satellites. But that's always going to be more expensive than other ways of providing coverage.
In a real case, I expect this to be part of a terrestrial, drone, and maybe even balloon network.
The article suggests the satellites would operate from LEO orbit which generally means a few hundreds to 2,000km above the planet. This is less than 10ms even accounting for ionospheric effects.
Bandwidth would be a huge problem, though.