OK, this thermal imaging stuff is all great.
This year, the late adopters will be using it for what the early adopters have been doing for years - equipment and heating inspections, spotting people and wildlife, looking at footprints fading, pouring coffee, watching yourself pee.
So: Thermal cameras are now cheap. Gesture recognition becomes easy. Occupancy sensors become easy. What's now possible? What have you wanted to do for ages, but has suddenly gone from a £2000 price tag to £20? £200 to £2?
Something in the cooker hood to both allow gestures to flick pages on a recipe book projected on the wall, and watch the pan contents temperature?
Something in the room thermostat, to watch skin temperature rather than thermostat temperature?
Bah, I dunno. This is a wonderful toy, at a new price. Let's avoid some gits getting patents on stuff that could be fun. What should be built?
I do kinda wonder if the mass market the manufacturers are hoping for is really there.
It's still (and likely to be forever) too expensive to come down to 'toy' prices, or even low enough to be built into a phone (though I have no doubt someone will, just for the publicity).
It will obviously go into the lower end of markets already using them - electrical/electronics/HVAC. I have little doubt that someone will make an industrial multimeter with a built-in thermal imager.
There will be some niche applications in occupancy sensing and maybe some safety applications and industrial condition monitoring.
And probably CCTV/security, though resolution is probably not good enough to do more than use as an initial detection system to target lighting & higher res visible cameras.
For gesture recognition, the ITAR framerate limit is a major obstacle - 9fps is just too slow. However Flir may be in a good position with the Lepton, as assuming it as some reasonable amount of built-in processing power, some of the recognition process can be done on-module at a high framerate, so they could potentially produce something that couldn't output an image, and so would be outside ITAR scope.
There's little doubt that phone thermal imagers will be one of this year's hot (!) Christmas toys, but I suspect the novelty will soon wear out, leaving a bunch of niches of various sizes and market values.
I also think that the Flir One was primarily meant to be a technology demonstrator, and their main eye is on niche markets rather then the consumer.
Even ignoring the spec differences, Seek Thermal is the obviously better way to do a phone add-on. For most people it will be an occasional-use tool, so reducing the hardware electronics and size, hence cost to absolute minimum makes far more sense than having an external battery and powerful processor etc. in the camera.
The Lepton module, with its internal processing is clearly aimed at applications where you don't have the sort of powerful processor that's available in a phone.