The death of XNA, does it matter?

No, not really I would say. I would however like to split XNA into two parts:

* The first being the XNA API
* The second being Microsoft implementation of that API.

What's really dead (not counting XBLIG) is Microsoft commitment to further develop the API and the implementation of it. The API in its current version still exists and MonoGame has made a great job implementing it and making it cross platform. In fact, I would say XNA (the API) is way better of this way since it's now an open source, cross platform product driven by the community around MonoGame.

So is it all happiness and flowers from now on? No, there are risks of this unoffical handover and that risk is fragmentation of the API. Or even worse, bad design decisions from the MonoGame team on where to take the XNA API in the future.

Let's hope that the MonoGame team keeps the quality of the work done so far and make sure to donate to the project!