The counterpoint to mesh is that it's a) much more complex to implement b) uses more power since nodes need to relay each others' communications.
It will be interesting to see how it plays out, and who ends up crying in whose martinis as you say.
Which is the catch. Run too many nodes at once, and the signal-to-noise ratio will rise and your signals will start to fail to decode. What this generally means is that nodes at the edge of the network - the ones with the lowest SNR budget - start to drop out first. I'd be interested to see what their link budget and range is with 1600 devices on the same node, or conversely how many devices they can fit on a node at the advertised link budget and range. (Also, just how well they cope with a mixture of near and far nodes.)
* Star networks are infrastructure heavy, requiring build-outs on part with cellular networks, and having low per-base density.
* LoRa's CSS (Chirp Spread Spectrum) approach doesn't promise to scale well, as devices start to interfere with each other over its very wide spectrum scattershot. There are other problems.
Disclaimer: I work for an IoT mesh company (SSNI/WiSun) and we're busy cranking out massive networks that actually work. Sadly you can't get a dev kit at Fry's yet.