Really not thinking like me.
Number one that example with a light blocker is not a ftir setup. Ftir setup has IR inside a bit of acrylic contact with the acrylic causes it to be release for the camera below to see.
Now camera can have a light filter on it. Issue you run into is in direct sun the surface of screen where IR blocking matterial ie the plastic surface you touch or the acrylic gets hot enough that it ends up emitting more IR than what touching the acrylic will release.
So effectively the camera ends up seeing a white out. Now finding a clear liquid that has good IR blocking property that you can keep cool that is not going to kill people would be kinda idea. If you still have to have IR blocking material what is used on car windows and the water under it just cooling for it that would be fine too.
Currently the silicone surfaces are using air as there liquid. Two major problems AIR is compress and expands a lot number 2 they don’t seal the edges so the air can get press out equaling stuck. Basically they are building a flat form of a suction cup and wondering why it is sticking from time to time. Depending on texture to make it un stick. So any defect in texture it will fail. Also as anyone who as worked with silicone knows it don’t keep it shape over time. So the textured shapes will deform off. So replacing the screen comes a really common event. Even just sealing the edges and having air presure under screen might fix some of problem.
Reason why I am suggesting this it might be the way out of the texturing problem. No need to texture would make ftir screen construction so much simpler and most likely last longer.
Also due to the silicone always being in contact with screen by some amount. I suspect the high points of the silicone are the reason why raw finger on acrylic can be brighter than silcone on top. Not all the light that should be staying in the ftir is. Basically its creating interference and reducing performance from being in contact all the time. Just raw finger on acrylic is temperamental.