OK, one more piece of spam.

Nokia had a number of options to get Maemo 5 on older devices:

Provide a one-off unofficial hacker edition

Much like it did for the 770, this gives existing users access to (most) of the new software at the expense of continued support past one or two releases and stability. While it would provide an easy way to get Maemo 5 on your N8x0, it's essentially money down the drain after just a few updates of the officially supported platform. It does not help get older hardware supported in new kernels, it does not provide many updated core libraries for old devices and it does not provide longterm support. It is good press, however.

Provide an officially supported Maemo 5

This is what most people are clamoring for without understanding all of the issues involved. On the surface, this sounds like an appealing solution. Current owners get the new hotness, everybody's happy, right? Unfortunately it's not that simple.

Providing an officially supported backport (unlike an unofficial hacker edition) brings with it a whole slew of QA requirements that increase costs by several orders of magnitude. This presents a problem for Maemo Devices, because the Maemo platform is only just coming into its own as a part of Nokia's core portfolio and they simply do have the personell or the money to support an official backport (especially at a crucial time like this that will make or break the platform).

An official backport is simply too expensive to be an option.

Support a community project

This is the option that really shines, even though most people wont understand why. Supporting a project like Mer has the highest return on investment. You get long-term support for your existing hardware, you get (most) of your new software on old hardware and you get a sane distribution that may provide a model to keep your platform viable in the long term (even though Nokia may not realize this one yet ). Mer ends up benefitting the whole of the open source community and not just your device owners.

So, given that there's a limited amount of both money and personell to put towards fixing this problem and given that providing official support likely excedes the limits of both, you're left picking between a on-off hacker edition and a community-supported backport that may be able to provide support for your existing users for a much longer period of time, which would you choose?