The author is 100% biased as the Chairman of York, one of the SDA's smallsat primes.
I agree with him for smallsat constellations. Planned obsolescence and short (5 yr) life spans make sense for cheap satellites that can keep getting replaced with upgraded hardware while the industry grows up.
I think there'll still be a small market for mission extension for $atellites, but that's a far cry from drawing an exponential line of LEO traffic and calling it a refuelling business model.