Yes it is, we get less than other countries, but it's still more than viable. That graph means nothing. What matters is does it pay for itself within its life span. The answer is yes it does.
And that's at around ~10 efficiency.
Hmm...I'm going to look at some figures. Feel free to tell me if any of my figures are wrong.
Solar power on the ground in the UK averages about 1000 KWh per square metre per year.
At 10% total efficiency, that's 100KWh per square metre per year.
The cost of generating electricity using conventional power stations is in the region of 8p/KWh. So 10% efficiency solar panels in the UK generate about £8 worth of electricity per square meter per year. Maybe £10, to be more generous.
That's not going to cover the cost of the cost of the system within its working life. Even if it remains working forever without any loss of efficiency, you'd probably be better off putting the money in savings instead.
I'll look at it from another angle:
Companies selling solar power kit and from advocates of it (i.e. I've chosen sources biased in favour of solar power in the UK) state that a fairly typical "3KW" setup (which will not supply 3KW) is generally stated to probably provide ~2500KWh per year in England.
That's maybe £250 worth of electricity.
The cost of having that system installed is about £12,000.
So it would take about 48 years to recover the cost. Assuming that the system worked for 48 years without any loss of efficiency at all (which is obviously not going to happen) and ignoring the loss of the interest that you could have got by putting the £12,000 in savings.
I am far from convinced that PV is anywhere close to paying for itself in its lifespan even in the south of England, let alone elsewhere in the UK. And that's with panels with efficiency far closer to 20% than the 10% figure you said was enough for PV to pay for itself here.
If you think I'm wrong, feel free to prove it. The key numbers are the amount of solar energy at the ground here, the efficiency of solar power systems commercially available today, the generating cost of electricity today and the cost of buying solar power systems today. I've stated the numbers I've used for those things. You have to prove that I'm out by at least a factor of 3.
Then there's the other issues that haven't been mentioned yet:
Solar power is wildly seasonal and variable, so even if we were able to generate enough from solar at some times, we'd still have to maintain other sources continually ready to pick up the slack when the weather is bad for solar, as well as every winter and, of course, every night.
Solar power from homes can't possibly supply a national grid regardless of how much it could generate. You can't run a national grid supplied by millions of very small-scale power stations none of which are under the control of the grid. Any kind of national grid requires
controllable electricity generate in order to match supply to demand.
Nor can it reliably provide the home with power 24/7, so you can't have everyone off the grid and generating their own electricity. You still need the grid.
Finally, there's the scale of power issue that I've already mentioned. Solar panels
might in a few decades drop so much in price that they're worth sticking all over the place and thus generate a lot of power by sheer scale. Nuclear fusion
might in a few decades reliably have a higher energy output than input and if that happens there will be a massive superabundance of electricity, which will make many things possible, e.g. cars powered by hydrogen fuel cells - if you can generate 20 times as much electricity as you need, you don't need to care how much you lose by splitting hydrogen off from water, transporting it around and then passing it through a fuel cell to generate far less electricity than you used to get the hydrogen in the first place.
I'm still standing by my original position - it is not certain that renewables will improve so dramatically in the next ~40 years that nuclear fusion would be pointless. In fact, it's extremely unlikely if not downright impossible.