back-of-envelope: conservation of momentum: 2 (num photons) h/lambda = (100 kg)*(c/3), so 5.9×10^36 photons needed to propel a 100 kg object to 1/3 speed of light. 1.5×10^18 joules of energy, which is 11% of the energy output of the US in 2001 (according to wolfram alpha). That's a 5.8 Tera Watt laser running for 3 days straight. I haven't even gotten into dispersion of a laser beam. The ludicrousness of this proposa…
If we're going to be doing space travel on time scales where decades don't come into the picture, we're going to need devices with this kind of energy. Access to that kind of energy isn't ridiculous if your infrastructure is already in space, so it's only ridiculous in our current context.
The notion of creating a web of railway networks across an entire country was completely impractical before the Bessemer steel process.
I haven't even gotten into dispersion of a laser beam.
Also, diffraction lenses of ridiculous size become practical in micro-gravity so diffraction isn't an issue either, so long as your infrastructure is already in space. This is precisely how Planetary Resources will eventually make their giga-fortunes; by pioneering entirely new forms of infrastructure for an entirely new context of civilization. (This new context may not have much to do with earthly civilization, or even with humans.)