Sure, so you see the contradiction on theory and reality? F22 and SU57?
There were old articles where LM and ATF people claimed -40 to -70 Dbsm “on a certain angle”, but, as far as we see from all simulations (not only on the F22), those hyper low numbers are on very minor arcs of the skin, like an arc less than a tenth of a degree – if they appear at all.
Then, most F22 simulations (most RCS simulations in general tbh) use very simplified models (nowhere near the detail as in the one posted). Suppose it is true, that RCS is super sensitive to shape, as some claim, and thus minor deviations from the original matter a lot, then most RCS simulations out there are unreliable simply because the models are not super accurate.
As you pointed, gaps and wedges are a reality, but, they also happen to be sources of diffraction. If shape is so important on those sorts of calculations, to the point even a 0.1º in wing sweep makes a difference, then those simplified models just don’t cut it. For comparison, these are some CAD renders from LM themselves, back in the 90s/00s:
As you can see, they’re also “perfect tolerance” models, and look kinda low in polygons (expected for computational power in the 90s). So, can we rely on RCS predictions done on models with the same quality as that? If shape is such a super important factor, my guess would be… no.
But that’s what we mostly see on simulations: things are perfectly air-tight, gaps are absolutely non-existant, and a bunch of other idealized conditions. I can kinda understand using PEC (perfect electrical conductor) for aircraft skin to try isolate shape effect alone, but is it good if shape is wrong and so sensitive to angles? Also, as EM are interferring waves, can you say for sure a specific patch of skin wasn’t deliberately made of metal just to cause wave cancellation from an echo from another part of the airframe, at certain conditions? It’s a lot of “ifs” for such seemingly important factors. Then, reality is 5th gens have composite skin – not gold, not silver, not copper, and certainly not PEC --, edge/wedge/tip treatment with RAM, RAS who knows where and who knows how advanced electronic means of interferring with incoming radar waves.
Then, with all that said, it’s hard to see the claimed -40 to -70db in simulations with airtight models, can we expect a real life model with manufacturing imperfections to achieve the same? Were those claimed numbers coming from their own simulations or from anechoic chamber testing with a 1:1 scale model? The information isn’t clear to that. Many of the reasons given for the SU57/J20 (at occasions even for the F35) lack of stealth can be seen for the F22 on images from that tour video (gaps, weird shapes, possible corner reflectors, etc).
So, again, I see those simulations as mere curiosities, I don’t think they are reliable, nor are the manufacturer numbers nor claims we see online.