He thinks that Asus made and sold somewhere between 4.5 and 4.6 Nexus 7 last year. I should not heckle Mr Evans like I have; his figure could be within a reasonable margin of the real figure (minus or minus 10%).
Mr Evans added up Asus' posted sales figures for q3 and q4 2012 (4.8 million) and then factored out his estimate for the tablets that Asus made that weren't the Nexus 7. As you can see, his upper limit assumes that Asus didn't sell significant numbers of any other tablet besides the Nexus 7, and we know that is almost impossible.
and now that I've looked at his methods I have to say that I think he's wrong. I think he didn't sufficiently take into account the other tablets that Asus sold in 2012. I have little to go on besides the past sales figures that Asus has released, but I think they sold more than he expected. For example, in q2 2012 Mr Evans has Asus charted at selling 900k tablets. That was before the Nexus 7, so wouldn't you think that Asus would sell at least that many of their non-Nexus 7 tablets in the 6 months after it is launched?
I do. In fact I would go so far as to bet that Asus sold only 3 million to 3.5 million Nexus 7. And if I am wrong then I will go double or nothing that my guess is too high. Asus is more likely to have sold under 3 million Nexus 7 than they are to have hit the 3.5 million mark.
Then again, I have no more accurate info than anyone else, so I could once again be completely off the mark. But it is fun to speculate.