I can see Silverlight perhaps finding a niche in internal application development - but that's a pretty big come down from the original goals of the technology (i.e. a Flash killer). So wouldn't that make Silverlight, if not dead, severely wounded?
They are simply refocusing their efforts to get Silverlight (which is a pretty good piece of engineering IMO) put to use elsewhere.
As far as I can see, that is a pretty good strategic decision.
Silverlight and Azure are different things, but they're both marketed towards developers who would like to take functionality, much of it in the form of internal apps, and make it accessible on the web or otherwise over WAN. I know this, because some Microsoft folks were pitching Azure to me just yesterday!
Yes, things don't make sense if you're stuck on "compete = comparable." A little reflection, and you should realize this isn't a very good assumption. (Teleconferencing and airlines, anyone?)