I regret buying the 950. Excellent hardware, but unstable battery hog of an OS, with an app situation that's bad and getting worse. WP 8.0 was the high mark. All the changes since then made things worse.
From the user point of view probably nothing.
You should follow this tweeter discussion from Sinofsky
https://twitter.com/stevesi/status/733699065763004416
It wasn't the difference between desktop and mobile stacks that pushed most away, rather the breaking changes between WP 7, WP 8, WP 8.1 and WP 10. As explained recently on the Ars Technica article,
http://arstechnica.com/information-technology/2016/05/onecor...
And as a developer I cannot really get why WP 10 is still so buggy (according to the reports) versus the already stable 8.1.
Microsoft eliminated QA as a job role and laid off all of its QA staff in 2014, with rather predictable consequences for both Windows 10 and WP 10. Plenty of articles about it if you Google around.
If anything, it's surprising that things have turned out as well for them as they have.
It does make sense, though. QA teams cost a lot of money, and don't produce anything at all. They're a cost center. The only thing they do is keep your quality up so that customers don't get mad and leave.
MS doesn't have to worry about that. People weren't buying WinPhone anyway, and no one's going to abandon Windows (PC OS) because of bugginess. So using the users as beta-testers makes perfect sense for them. If you don't like it, then don't buy MS products.
Remember any of these: Pocket PC, Windows CE, Microsoft Kin, Windows Embedded Handheld, ....
MS spent $1 billion for Kin alone. When they consider market to be important to their future, they'll keep going and spending money. Setbacks destroy their partners but they just wait and try again when time is right.
MS is still mainly business and OEM sales company. Consumer-centric divisions (including XBox, Windows Phone, Bing) are irrelevant from a profit standpoint. MS is perceptually trying to get into mobile because they are fearing for their business and OEM sales if they don't have the presense in mobile.
Useless product, nobody uses it, but the fact that they kept investing in it made most competitors lose their nerve.
And Myerson has the audacity to say "We always take care of our customers, Windows phones are no exception. [...] regardless of a person's phone choice, we want everyone to be able to experience what Microsoft has to offer them."
For me, the appeal as hobby development was that I wanted to support a third ecosystem, have my love-hate relationship on the MS stack back to the MS-DOS 3.3 days and I love what they are doing with C++/CX (VC++ finally catches up with C++ Builder) and the whole AOT compilation for .NET (MDIL and .NET Native).
But with this scenario there isn't much value in building up WP skills specially since WP10 is still quite buggy, even as an hobby.
Sad, because WP is actually much better from tooling and architecture point of view, but the way the whole WP 7, WP 8, WP 8.1 and the yet to become stable WP 10, just drove everyone off.
Personally I was kind of pissed off, when the 512 MB devices were left out of the last preview update.
Also it was the only viable alternative to Android, on the countries where iOS devices are seen as too expensive, specially the ones where devices aren't subsidized.