Anything that increases the time and effort required for both detection and tracking is worth it, and all else being equal a LO aircraft will always be more survivable in a dense A2/AD environment than one without LO features. It's a continuum, not a binary situation where stealth is either perfect or defeated.

The USAF was well aware that it would not be able to preserve the impunity with which F-117s flew over Baghdad in 1991, but that was never the objective.

Shaping remains by far the most important factor for reducing radar cross section, followed by the use of radar-absorbent materials (RAM) in the airframe and the application of a RAM coating on the outside.

The F-35, for instance, gets most of its RCS reduction from shaping and the composition of its fuselage, and was designed with extremely low tolerances in order to rely a lot less on needing a top coating to cover seals, rivets, etc than earlier LO aircraft. So even when the LO coating degrades, as it does on long operational deployments, it doesn't catastrophically increase the F-35's RCS.

That's an example of the sort of incremental improvement I'm referring to. The F-35 has come a long way in terms of LO from earlier US aircraft. Compared to them, it needs a lot less RAM coating, it can use less hazardous materials for that coating meaning you don't need as much specialist equipment to apply it, and it's more resilient to materials degradation. Those are all things it's somewhat difficult for even peer competitors like China to catch up to.

I should clarify comments was with respect to F35. I'm not disputing the importance of stealth, but depending on how far stealth detection has progressed, stealth may be merely table stakes. We know where stealth sits on the continuum relative to none stealth, but not relative to modern anti stealth, i.e. LO/VLO may be substantially less survivable vs modern sensors / detection methods vs 20 years ago. Question is, has incremental improvements in coating extended the gap, maintained it, or diminished relative to detection. Doesn't mean stealth is obsolescent (if anything it would become MORE necessary), but not standing still doesn't mean not getting lapped by detection/tracking/targetting potentially improving at faster pace.

When I say LO emphasis materials over shaping, with respect to F35, I mean F35 is not an opmital shaped flying wing for stealth. It has optimized shaping for what it is - an ellipitcal wing designed for multi role. It's compromised shaping relative to say, a B21, or renders of NGAD (both flying wings), and likely further fat amy belly compromise to accomodate S/VTOL (vs F22 or even J35 geometry), hence F35 has suboptimal shaping relative to optimal stealth planform. My understanding is coating can reduce RCS by another 30% on top of geometry, ~10% less detection range (inverse fourth power law), which is not catastrophic if lost. But at same time, math suggest incremental coating advancements, which F35 is stuck with to improve stealth, since geometry is locked, means it won't be game changing, it's merely what there is. Hence I'm not convinced incremental coating improvement is enough. IMO pretty telling most 6th gen proposals (even though renders) are all flying wing or blended delta wings.

And as much as F35 coating is improved relative to F22, this report suggests it's still maintenance nightmare. As for whether coatings are difficult to catchup, PRC has got very good material science in last 10 years. It seems more easy to iterate than engines.