Oh yes, there's all sorts of problems, I know. I think that ordering several hundred billion panels all at once might ... ahem ... cause a teensy-weensy glitch in the world market for example.
But suppose we had (say) 30GW+ nuclear baseload (that does roughly cover minimum use ~4am), eg see http://www.leonardo-energy.org/drupal/node/2423 and another 40GW+ of solar, and decent tidal and wind, that probably could meet all UK power consumption just about during the summer, at nominally zero carbon, especially if we had time-of-day "smart" metering to very strongly encourage people and businesses to move their usage into the daytime when the sun is out and/or the tide is running and the wind is blowing.
You keep some decent (and efficient) natural gas plant that can be cranked up and down very fast to cope with short-term mismatches in supply and demand, but charge a huge premium for their power as a carbin tax.
And then in the winter maybe we could burn recently 'stored' solar energy, eg from managed woodland and other biofuels. Again, nominally zero carbon, with backup from gas which is the least-bad (least-carbon-intense) fossil fuel source.
I think that it's actually plausible if it was thought to be the right thing to do, for whatever reasons (global warming, avoiding funding dodgy oil producing nations such as Norway^W^W^W B^> )
Incidentally, some of the nuclear baseload can be done away with and replaced with RE as baseload if we do any or all of a number of things, eg better/bigger interconnectors to other countries with different load/sun/wind profiles to us, storage improvements such as (distributed redox flow batteries or melted-salt heat stores, etc, etc... (We only have a seedy 2GW of interconnector from the UK to France right now, with another one being built to the Netherlands. Others have been proposed.)
Rgds
Damon