Was thinking it would be nice if it slowly rotated left & right when it burned.  It would only need to actuate 1 chopstick.  It would provide some more entertainment but take more room.

Went with a full orthogrid behind the most melted parts.  Either way, the melting seems reasonably solved.  Noted it's a very gradual melting which exponentially expands after many burns.  It's always going to gradually melt, but the stringer patterns seem to have extended the lifetime just enough.  The heat break & the masking tape are other areas which are going to limit the lifetime.

It completely dropped the 1st layer of bridging in the nose but managed to do better in the cylinder.

The front was sitting a bit proudly.

The nose held up better but the cylinder was unchanged from just the vertical stringers. 

The mane problem was shrinking of the catch point area.  That had shrinkage on the edges.

Thinking epoxy channels on the edges should fix the mane trouble spots while the nose is already fixed.  The common reason for the melting is the large number of small windows.  If they were replaced by big holes & epoxy was applied all around, it wouldn't melt.

Lavender & full moon are the only good ones, locally.

The chopsticks have deformed to the point of making it harder to empty the ashes.  They might become PETG with a long weld for the hinge.  Stuff which melts as slowly as that might be easier to just melt back into position.

Surprised there isn't a single part adhesive in filament form.  The printer would extrude it & it would cure into something heat resistant.  There is no UV cured filament.  Resins have higher melting points but are expensive.