Problem 1: Spotify isn't profitable.
Microsoft should only acquire Spotify if Microsoft has some way to make it profitable. It's *possible* that with a distributed cloud service like Azure, Microsoft running the streaming "at cost" for Spotify could turn it profitable, but thats very, very unlikely. If you look into what is keeping Spotify in the red, its the heavy weight their free users on their operations. Microsoft probably can't do much to fix that.
Problem 2: Spotify has competition that may be better suited to acquisition.
Pandora on the other hand, while it has a smaller userbase, has a much larger "install base" than Spotify, and is profitable because they have a better subscription to user ratio. Smaller userbase also likely means it would cost less to acquire them.
Theres also pros and cons to the two services, most obviously, Pandora's "Radio" feature is awesome thanks to music genome while Spotify's constantly plays the cheapest songs and makes no real effort to "learn what you like"; but Spotify also has a much, much larger library to draw from for users who like to make their own playlists, while Pandora users can only make playlists if they subscribe, and are limited to Pandoras smaller library.
Problem 3: "Whats in it for Microsoft?"
Seriously. What does Microsoft gain from owning Spotify (or a competitor like Pandora thats actually profitable)? Currently, you can use Spotify or Pandora on Xbox, Microsoft Store PC app, or the web. Generally, a purchase like this would be made to secure support for a currently unsupported platform, or to block support of a competitors platform (eg the reasons why Google buying Waze were to kill off Microsoft support and block Apple from having an easy path forward in navigation).
Considering Microsoft is supported by Spotify on every platform, and considering there is no platform where "platform exclusivity" would help Microsoft (no one will drop PS4 or MacOS just because Spotify [or any other music service] became exclusive), I dont really see any reason at all for Microsoft to gobble up a music company. Also keep in mind that killing off support for competitors is somewhat against the current ethos of "go where the customers are".
The only scenario where owning a music service makes sense is one where Microsoft needs to ensure they have one (getting back into mobile or something). And I dont think that is a realistic scenario.