It may be a year old, but the arguments @vectorselector is raising are pretty valid, IMHO. I am also quite puzzled about the current situation, given that the key maintainers of mup(x) and cluster are not part of the community anymore.
The solutions Arunoda and his team offered were just awesome, and they worked together more or less out-of-the-box. I get the strong feeling that Meteor’s fragmentation starts to get to a point where it’s a huge number of puzzle-pieces I have to put together, just in good old MEAN stack. I doubt this will really bring Meteor forward.
And yes, I know that I might as well use Galaxy for hosting, but the fragmentation goes far beyond pure hosting. I had lots of a-ha moments just recently when I tried to achieve something and learned that this would require me to replace a vital part of my stack.
