The problem with that graph, assuming an R rate of 4, slipping to 3.6 ish by lockdown then 0.7 ish until end April is the maths.
Start date 16/1 with 1 person and an R of 4, 85 would have been infected by end Jan. By end Feb that would be 350,000 and by mid March we'd be up to 19.5 million.
(Didn't Oxford say 50% of the population?)
Then if we taper to 2.5 and 0.7 over the next 10 day period and remain at 0.7 until end April the sum total of infections would be 134 million by the end of April.
Now in all honesty, I don't think the graph is far away. We probably did have huge spread without knowing it.
But if you present this graph from Imperial and this profile of R=4 up to 10/3 approx. then dropping to R=0.7 by 23/3 approx then you also have to accept that somehow we have herd immunity now, regardless of what the anti-body tests are coming up with.