Slightly longer version:
There are basically 3 different aspects on this:
1) Static/nominal DB values
2) Dynamically (in-sim altered) values
3) Altering the IRST (and visual) model to curtail the "volume search" capability
1) The DB values remain to be revised. We will probably need to make some pretty significant tweaks to the signature modifiers in order to incorporate these factors. As it stands, the current modifiers do not offer the flexibility necessary for this.
These are the modifiers currently in use:
2) In some cases the visual & IR modifiers are indeed changed in-sim. An example of this is supersonic/hypersonic skin friction. You can see this clearly in action if, for example, you attempt IR detections (preferably head-on or side-on) on an SR-71 first at subsonic speed and then at M3+. In the latter case you should observe a substantial increase in detection range with a modern IRST.
3) The biggest handicap of the current model IMHO is not the signature details, but that fact that visual/IR sensors are allowed to function as mass volume-search sensors, which IRL is possible only at far-less-than-maximum ranges. With the exception of specialized "staring" sensors like DAS, most such sensors suffer from the "soda straw effect" which limits both their volume-search ability as well as their capacity to keep track of existing contacts. (Abhirup Sengupta describes these limitations decently here: https://www.quora.com/Can-5th-generatio ... Sengupta-5 . He also mentions some of the IR signature suppression techniques & modifiers that you refer to).
At the most recent CUC we presented our preliminary plans for addressing this. Here are a couple slides from the presentation:
I think getting this implemented will go a long way towards balancing out the pros & cons of IR systems, and combined with suitable changes in the DB modifiers can rectify the issues described above.
