There's what papflesje just said, plus the fact that those estimates don't show us which exchange rates they used for the estimates, and that's a huge factor these days.

In a week Sony will release their earnings and we'll probably be able to see how right those estimates are (I would bet they're wrong).

@ NJ: what would your guess be? Even higher, or closer to the second link that mrstickball posted?

I think these estimates might well be correct at the pre-currency-crash levels, since Sony was expecting to break even on hardware towards the end of the fiscal year. Right now it depends a lot on the mix of currencies Sony uses to pay manufacturing/components, so it's really anyone's guess but it can only be higher than $448.

How can we see if the estimates are on or off track?

Even if they have the PS3 hardware revenue as a standalone number (I have no idea, which is why I am saying like this), the fluctuations in currency and continues drop in price for the PS3 hardware, should make it impossible to make anything more than another qualified guess.

Unless they state the current cost of the PS3..?

I know there has been some questions posed in regards to iSupply's estimation methods for Wii production costs so I wouldn't be surprised if the same issues were carried over here as well. In the past we have seen:

actual cost = O(iSupply estimate)

In short, iSupply's estimate is probably a good upper limit estimate, but probably not a very good ideal estimate (Ω) or realistic estimate(Θ).

There are a lot of difference between Wii and PS3 production though, so who knows.

I think it is all manufactured in China, but of course there is another question of where as the components come from

Because PS3 hardware is probably the only big thing they're losing money on. If we see them losing hundreds of millions throughout the quarter, it's very likely they're still losing a lot of money on hardware (at least it would look like the only plausible explanation).

For example, losing $500 million at 4 milion PS3s shipped would mean at the very least they're losing $125 per unit according to these assumptions.

That's a pretty bold statement..I'd have to respectfully disagree.  But, as you said, it's anyones guess...that's your guess, and this is mine.  But I don't think it's at or over $448.

OK I should have been more clear. That sentence I wrote is under the assumption that iSuppli was correct at the moment I mentioned.

If you are correct, you're not expecting SCE to post a loss in either of the last two quarters, right?

Ah, that way. Although it would still be pretty bad method for indicating whether or not their guess is on or off-track, as the production cost will be much higher than in the beginning of the fiscal year than at the end of it.