We know in 2010, the year before Elop sabotaged Symbian, NOKIA sold 103.6 Million smart devices.

We know (because ZTE have announced it) that the cost of a license for WP7 is $24 to $32 depending on the standard of device it's to be used on. To be conservative we'll use the lower figure of $24

We don't know the exact number of developers NOKIA employed to work on Symbian, articles I've read seem to vary in estimate between 3,000 and 6,000. To be conservative we'll use the higher figure of 6,000.

So let's take NOKIA's unit shipments in smart devices in 2010 (i.e. before Elop's act of sabotage) and multiply it by the license fee per device that NOKIA would have had to pay if those devices had been running WP7 (we'll have to use our imaginations a little here as there's never seemed any likelihood of WP7 devices selling in these vast quantities):

103,600,000 * $24 = $2,486,400,000

Let's then divide that by the number of Symbian developers NOKIA employs so we get an annual cost per developer:

$2,486,400,000 / 6,000 = $414,400

So NOKIA would only make a saving from this plan of action if the average annual cost of employing each Symbian developer was greater than $414,000.

Mmm... do you think that's likely?

And we were being conservative! If the number employed in the Symbian team was at the lower end of the spectrum and the license fee is at the upper end then we would only be talking about making a saving if the average annual cost of employing each Symbian developer was greater than $1,105,067 per annum!

Then of course you need to start factoring in the increased hardware costs because WP7 is much more demanding on resources than Symbian and it supports a much narrower range of components.

You also have to consider the loss of control, M$ developers are not directly employed by NOKIA and so their priorities will not always be the same as NOKIA's priorities.

It seems very clear to me that every single thing Elop has done has been to the benefit of M$ and to the detriment of NOKIA.