One feature/bug of this kind of technology is there is a secondary "height over bore" issue. The thermal camera and the red dot have different entrance pupils, with the red dot behaving like a normal red dot, but the thermal camera pupil being above the red dot pupil.
I think a more ideal technology would be to have one big entrance pupil, but use some sort of mirror to create the alternative pathway for the thermal radiation. Perhaps that technology doesn't exist at present, or is impractically expensive.
But because you have two entrance pupils and only one exit pupil, depending on what distance you focus at, the thermal overlay can appear a little mis-matched from the direct optical i.e. red-dot pathway image. So if you have the two images aligned at 25yds, then when focusing at 5 yds the thermal overlay will look about a quarter inch off (at the point of focus). There's a good amount of adjustment available on the unit, but around 5 or 6 yards appears to be the closest you can get the two images to align before you max-out the adjustments.
In practice this means your thermal overlay "drifts" a little from the point of aim, depending on what distance you focus at. This is much like using hold-over on a regular scope, so it's not a new issue, but just one more of a familiar kind of issue people deal with when using optics.
Future iterations of this kind of device could easily eliminate this issue by incorporating an in-unit rangefinder that adjusted the overlay according to the distance to target.