I'm still kind of puzzled why so many of you are so sceptical about the iPad 3 getting a screen with double the resolution in both directions. The only justification for this scepticism I see is: "it's too expensive", "the GPU would have to push 4x the pixels", "it would use too much power" and "nobody can produce enough of these screens fast enough".
None of these arguments seem to be based on more than just assumptions of what is and isn't possible, and history has proven that sometimes, Apple will pull out a rabbit from the hat, that proves more things are possible than you would have imagined beforehand. Such as a 960x640 iPhone screen, a tablet with 10+ hour battery life, or an ultraportable that's less than an inch thick.
Let's run through the arguments AGAINST doubling the iPad 3 resolution:
"It's too expensive"
I think this argument is bollocks, if you look at the price difference between the iPad 1 screens at launch, and replacement parts now, these screens now sell for 3 times less than they did a year ago. Produce stuff like this in the quantities that Apple sells in iPads will make any screen cheap enough over time. Even if a true retina-display is 3x as expensive as the current iPad screen, Apple would still be able to sell iPad 3's without taking a loss, they would just have to cut their profits initially, maybe to the point that they hardly make any money at first. It's exactly what they did with the iPad 1 to gain marketshare, and seeing how well that strategy paid off, don't be surprised if they pull a trick like that again. Then there's of course always the option to introduce and iPad 3 alongside an 'iPad 3 HD', the latter $100 more expensive, but with a higher clocked CPU and GPU and a retina screen.
"The GPU would have to be so insanely fast with 4x the pixels, it's impossible"
This is only true if you assume applications (especially games) can only render at the native resolution, which -again- is contrary to what history has told us with the iPhone 4 screen: applications can still choose the baseline resolution instead of the 2x resolution, and all of the development tooling is made to deal with this almost automatically. Developers could simply program graphics intensive applications at half-resolution, and native UIKit applications at full resolution. Apple could even adapt iOS to make the output resolution view-dependent: ie: an application that combines standard UIKit elements with custom-drawn elements could have all UIKit views at full resolution, and the custom-draw views at half resolution. The GPU could scale up half-resolution views at almost no cost and without introducing scaling artefacts (since that's what GPU's do all the time). For simple 2D user interface rendering, even the current iPad 2 hardware would be more than sufficient to push 2048x1536, drawing stuff like the iOS user interface is a piece of cake compared to a game like e.g. Infinity Field, in terms of GPU load.
"A 2048x1536 screen would use too much power"
AFAIK, most of the power used by LCD screens goes to the backlight, and adding more pixels doesn't make much of a difference compared to it. Sure enough a retina screen would use more power, but battery technology is also not standing still. If battery technology gets a little better, the iPad 3 screen uses a little more power, and a single charge would get you 9 hours on the iPad 3, instead of 10 on the iPad 2, the better screen would still be more than worthwhile.
"No one can produce these screen fast enough"
Why not? Is there any reliable indication that it is somehow impossible to produce screens the same size but 2x the resolution, at the same rate? How come Apple managed to double the iPhone 4 resolution out of nowhere, long before other phones had screens like that? And who else is using 2048x1536 screens? If Apple is the only customer, and they make big upfront payments for long-term screen production contracts, it would be the perfect opportunity for display manufacturers to invest in a complete new production line dedicated exclusively to producing ultra high-res 10" screens.
The arguments IN FAVOR of adding a retina screen to the iPad 3 are very obvious: it would propel the iPad 3 so far in front of competing products, that it would again be out of reach for a whole year. Nobody has a tablet with a screen like that on their roadmap, and nobody would be able to source enough of these screens at a low enough price, soon enough to have a competitive product out before the iPad 4 rumors would start. A retina screen would be a unique selling point for the iPad, it would allow Apple to better market it as an e-reader with a pixel density higher than print, it would allow for much more interesting applications of the HDMI output, and most importantly, it would be yet another huge blow to the competition.
In the boxing world, if you go down 3 times, you're out of the match with a technical KO. We've had the iPad 1, which exceeded everyones expectations, especially on price. Competition went down for a full year for the first time. Then came around the iPad 2, which handily demoted the supposedly superior Tegra 2 chip that everyone was hyping up, by humiliating it on its own game (graphics performance). Competition is currently down and doesn't really seem to be able to pick themselves up for now, there's still nobody selling any competing tablets in any meaningful quantities, let alone making fat profits off of them. Now imagine iPad 3 having a retina display, that would be the perfect TKO. Especially when the competition realizes they will not be able to include screens like that in large quantities fast enough.
Maybe I'm just being too optimistic, but hasn't Apple shown time and time again that they love to pick one signature feature for a product revision (be it screen, size & weight, a new port, a new UI, a new graphics chip) and then go all-out trying to make that one feature stand out so far above any competing product, that the presence of that feature alone will make people forget that the product might be lacking on other features?
I'm not an analyst, not even remotely so, but I would not hesitate to wager a crate of beer that the iPad 3 WILL have a retina display. I'm almost 100% sure of this. I'm also almost 100% sure that it will NOT launch in 2011, for the simple fact that it doesn't make ANY sense whatsoever. It's just a stupid rumor that went on to get a life of its own.