Pretty much this. Lack of interest/willingness to invest aggressively from the CEO, even though Maemo looked quite promising when announced.

Meego was a distraction and pretty much a mistake. I think it was mainly pushed by Intel so they have a reason to sell Atom. People may not remember but before "netbooks" Atom was supposed to go into Mobile Internet Devices (MIDs), which were much smaller than netbooks but larger than smartphones.

Netbooks pretty much came out as a "trial and error" device from PC manufacturers, who were failing with MIDs. If I remember right, Microsoft also had to extend selling new XP licenses post-Windows 7 launch, because of the netbook popularity, and because it didn't want Linux to become popular on them.

I think that's a bit unfair. Meego was scheduled to use Wayland instead of X. While I don't remember if it did, Sailfish definitely does. Maemo was also quite of an early adopter of Pulseaudio. Thus, the whole Maemo-Meego-Sailfish saga has been one of the pioneers in changing the stack SysV/X/ALSA -> Systemd/Wayland/Pulseaudio and friends.

The N9 was a really polished product. Offline navigation was incredibly good. I still use it. With a bit of care it could have been a nice product.

Was it bad enough to abandon all investments made into its development and jump to the less successful competing platform instead of incrementally improving MeeGo/Maemo?

I don't have experience with Maemo, but Meego was not particularly slow or awkward beyond being a young, immature mobile OS. In fact my N9 was remarkably stable considering the circumstances of its birth. And in fact I know several non-tech people who really liked their N9's, minimal ecosystem notwithstanding, so it had potential beyond a hardcore geek audience.

I had an N9. It was snappier than your average Android phone.

Didn't own one but a friend had an N900. It seemed like a few things still needed polish, but it didn't seem slow to me - quite the opposite. He hung on to it for a good while too.

My plan was to buy the next one at next contract change - of course that wasn't going to be possible. (I've always had a liking for the Communicator layout, going all the way back to Psion 3s)

When they announced the move to Windows Phone, I thought it seemed like a good idea. Meego clearly wasn't ready, Symbian was an antique, and Windows Phone was promising in so many ways -- even now I believe we will never see another smartphone OS as well-designed as Windows Phone 7.

Then some months later, for some reason, I ended up owning a phone with the last release of Symbian on it (Symbian Belle). And I realised that the original Nokia plan wasn't as stupid as it seemed.

Symbian Belle was a surprisingly smooth and pleasant OS -- much smoother in many contexts than Android at the time. It had some serious pitfalls, but it turned out there was quite a lot of productive turd-polishing that could be done. A lot of the good stuff on Belle was down to Qt Quick, which was the same framework as they were intending to carry through for Meego developers. And although I never used Meego, I can believe that it could have worked very nicely in the end, and a polished Symbian could have seen it through for a while.

But I was just as surprised to find out how much infrastructure there was behind it all. Nokia had an app store and billing platform serving a lot of countries and languages, that could bill you for apps either from credit card or straight from your carrier balance. They had one of the best mapping providers, a decent weather service, and a fine music provider. They had first-class hardware and a lot of public goodwill.

The experience changed my mind completely. Nokia could have done it with Symbian and Meego.