Picky
Not really following your argument here. But caution, I may just be being dense.
"The LH half of the superimposition shows the ablated material ahead of and behind the left wing, while the RH half again suggests to me that much of the leading edge of the left wing-cuff may be missing. So I am inclined to think now (as suggested by DrSyn for different reasons) that burn-through may have been initially far ahead of the wing and in particular much further forward than where the sweep back of the leading edge changes, and subsequently into the wheel well possibly from inside the wing cuff or fuselage.
I gather from this that you disagree with the proposition that the jagged projection ahead of the left wing might be the hyper-heated plasma bow-wave (many orders of magnitude hotter than anything ever seen before) associated with a failed and shattered RCC section 6 or 7 (see below). A knowledgeable friend of mine says that "Yes, it would be opaque as far as a camera is concerned, just like the heat off the desert sands creates mirages and heat-haze - but only much more so" That's where I got the idea of the "flat-plate" effect. Consider that the reason why the Orbiter's wing leading edges are so relatively blunt is because the designer's choice was to absorb high mach re-entry heat in the bow-wave created ahead of a blunt wing section. Once you take away
just one of those curved LE RCC sections, you are left with the inconel support brackets and an oblique flat plate of the aluminium wing. An erosion process quickly starts courtesy of the hyper-heated plasma and this is now being referred to as the "zipper effect" (which causes adjacent RCC tiles to be shed progressively out along the wing).
The trail behind (in that photo) is the detritus of that process. So as well as the inside of the wing being"eaten from the inside out" (once the gap between the flat-plate LE and wheel-well was broached), the LE flat-plate effect would have grown in length, creating greater adverse yaw and roll effects. Evidence for this has been found in the fact that the Orbiter parts recovered have had a fine coating of black aluminium oxide and the leading edges of ceramic tiles have sprinkled into them the once molten orange beads of one of the alloy metals found in Inconel 718 and 600 (the melted RCC support brackets). Hope you're following this line of thought. I'd be interested in why it might be invalid.
Basically it comes down to what I've said before. Give one of those solid RCC sections a square-on icy hit and shatter it and then there's really nothing retaining the pieces in situ (in addition to the all-important RCC anti-oxidant coating being pierced). After a few hot/cold cycles facing towards or away from the sun in orbit and the mildest manoeuvre would throw a large chunk of that shattered RCC tile (which is exactly what apparently happened from the manoeuvre logs - as recorded by Radar). The scene was then set for the erosion, wheel-well broaching and L.E. unzippering process. The early pre-dawn pyrotechnics over California would have been those section 6 RCC carbon fragments hyper-heating to a white glow and detaching.
A wedge-shaped elastomeric sacrificial Leading edge protective launch shield (for half a wing-span each side) is what the NASA Doctor is likely to order. That should adequately protect the RCC and burn away promptly without drama on re-entry, but without compromising its "in atmosphere" abort glide performance. In addition I cannot see NASA ever again foregoing in-orbit inspections of critical areas. Two-part exothermic mix "once only" overlays could be applied to any areas thought dicy. With the right composition goo, they just shrink into position and would protect against the sort of development that, with Columbia, started right at the entry interface and continued for 10-15 minutes.