I started thinking about this entry ages ago when I had an epiphany and I felt that it could be proven that a speed/price/throughput advantage would exist for "fibre" over "wireless". (Conversely, a flexibility and ability to function when the grid is "down", advantage would always exist the other way).
The basis of the proof is basically that any imaginable technology that improves speed over wireless electromagnetic radiation, would be applicable to wired electromagnetic radiation in the same proportion of improvement.
In the case of broadband, at one extreme is Fibre to the Home and the other extreme is Satellite broadband. Wi-Fi is closer to the FTTH side, then there is 3G style wireless broadband that relies on the mobile network which is connected via mainly fibre, and satellite, which can work even if the whole countries internet is down, theoretically. This doesn't really prove that FTTH is worth pursuing, because it is a tenable argument (although I would dispute it) that wireless will improve to the point that it is fast and cheap enough for everything we find important.
Back in Uni, it was proposed that solid state storage improvements were happening faster than hard drive improvements and could overtake them, and although a USB stick is enough for most things, portable hard drives still hold more and are used a lot, the storage advantage of hard drives is constant, due to the solid state improvements being equally applicable to hard drive storage.
Similar grid vs grid independent comparisons are common - road vs helicopter, grid power vs home generators, rail vs aeroplane, piped gas vs cylinders, water tanks vs dam & pipes, private dam storages vs large scale dams for irrigation.
With this in mind, simple goal based arithmetic could decide the balance between grid based and grid independent systems.
For Photo Voltaic electricity generation, I think we have got it backwards, which is why grid parity is not served well with thousands (or millions) of individual installations. If a household is after energy independence, a PV system with a significant amount of battery storage (batteries suitable for an electric car?) is actually very useful and would be a boon for extended power outages (like after Yasi). Because home PV installations are optimised to feed power back into the grid, they are quite useless as independent power sources, compared to diesel generators.
However, if we want PV power to compete with Coal or Gas, we need economies of scale of large scale PV based power stations.