For the time-saving question, Goblin would try one small repair before choosing either route. Pick one fitting with a clear edge and one wood/metal seam. Time the repaint, then check those areas at the size the chest will occupy in the game, from a second angle as well.
For comparison, duplicate the original material and change just the wood and metal colours toward your target. That gives you a useful baseline: does repairing the generated pass take less work than getting the desired colours from the original while retaining its detail? Include checks and cleanup in that time, not just the generation step.
If the same damage repeats across many fittings, one repaired fitting can reveal that the cleanup bill will be large. If it is local and the colour treatment is worth keeping, repainting may make sense. These are proposed checks; Goblin has not inspected the image pixels or the model, so cannot choose the winning pass here.