My calculus, best case scenario based on the December quarter ending on 3 Jan 2015, which would be the benchmark date for AMD's nearly 30 million statement.
AMD is estimating less than 30 million PS4bones shipped as of 3 January. This estimate would be based on it's 100% accurate knowledge of number of console APUs shipped to each company as of the same date (i.e. it's financial reporting period). This means AMD knows with a fair degree of accuracy, because of how supply lines work, that number of APUs shipped - APUs sitting at factories waiting to be turned into working consoles - working consoles in Sony/MS warehouses waiting to be shipped means <30 million consoles shipped to retail.
That's best case with Sony having low head room with shipments. If Sony has more headroom with shipments (i.e. 19.5 million shipped) that does drop Xb one sell through to ~9.5 million.
Personally I think <10 million sell through at 3 Jan 2014 represents a failure on MS's part, because at pretty much exactly the same time on market 360 was dead on 10 million. If Xb one has fallen behind 360 at this point in time things are only going to get worse for Xb one, relative to it's immediate predecessor.
PS4 has sold through 18.5 million (estimated) as of 3 January. There's no information to suggest this is not a reasonably accurate figure so we accept it at face value. This means PS4 shipped a minimum of 19 million, which would represents supply constraints in some markets, but probably about 2 week's sales for week 1 and 2 of January.
We have no officially accurate Xb one shipment or sell through information other than VGC, and AMD's "nearly 30 million" minus Sony's minimum 19 million shipped.
Nearly 30 million - minimum 19 million PS4s shipped = nearly 11 million max Xb one's shipped.
Xb one does not appear to be at all supply constrained, but then expectations for Xb one retail sales is considerably lower than PS4, so while 500K on shelves for PS4 somewhat suggests supply constraint, 500K on shelves for Xb one suggests reasonable supplies.
So assuming minimum 500K on retailer shelves, Xb one sell through as at 3 Jan 2014 is likely to be "nearly" 10.5 million or less. It may be as low as 9.5 million if there is nearly 1.5 million Xb one's on shelves. But I think 1.5 million in retailer channels is too high.
So my guess for Xb one sell through LTD is between nearly 10 million and nearly 10.5 million. Which means VGC over track of between slightly over 500K and slightly over 1 million.
“The fundamental cause of the trouble is that in the modern world the stupid are cocksure while the intelligent are full of doubt.” - Bertrand Russell
"When the power of love overcomes the love of power, the world will know peace."
Jimi Hendrix