I think even if you give feedback now, the units are probably already made and assembled and they can't do much to change the FR.
The TRI I3 really is quite coherent for a mishmash of 1 DD + 1 planar + 1 BA, very unique set in my collection, I still use it now and then. The planar mids are really lush and thick when amped. It has a soundstage in width/height/depth approaching earbuds, when amped, truly quite a big soundstage at the $100 region, but I suspect this had to do with the huge shells. So if the Pro has a smaller shell, I wonder how the soundstage will be affected.
Indeed, the original TRI I3 had a 3 kHz peak indeed, it could be jarring in poorly recorded material or with louder volumes (fletcher munson curve), but I'm sensitive to this area and by and large it is still okay for me. The 8 kHz area you see above is not too apparent in actual listening, I think that is a resonance artefact coupler peak on Crin's graph. The TRI I3 actually has a rolled off treble approaching dark treble, and is very suited for treble sensitive folks.
As for the 1 kHz dip, that's perhaps the crossover point or where the BA takes over from the planar driver. But in practice I think the planar may actually span a wide range of frequencies and the 1 kHz dip is not heard in listening. Perhaps planar graphs can't be interpreted the same as traditional DD/BA graphs? The planar mids as discussed above, are really thick and lush in the TRI I3, even at the 1 kHz area.