I am not an expert on cave sedimentology but I guess it depends on the type of cave/sedimentation and if it was frequented by humans. Charcoal particles in lakes are deposited through the air and by fluvial input so I guess that this could also be applied in caves with sediment accumulation coming from the outside by the same ways. However, I don't think this would work on a cave that was inhabited by humans that used fire.
I agree that human occupation would invalidate cave sediment charcoal analysis.
Cave sedimentation, at least in caves that I’ve worked in, appears to be similar to lake sedimentation though the majority of clastic sediments are deposited by aeolian processes (wind-blown for sand sized grains and dust-fall for silt and smaller particles) and and other sediment washed in, though not to the same extent as in lakes. One concern is that the sediments deposited in caves are frequently eroded (by wind or water) from sediment surrounding the cave entrance which may contain charcoal. In addition, I have observed structures that look like laterite but in thin section are seen to be charcoal grains coated with sand and clay. Clearly these factors add noise to the signal. If these factors also affect lake sedimentation can the charcoal record from either source be considered reliable?
Certainly, the results of analyses of charcoal frequencies in cave sediments can in no way be compared to those from lakes. This discussion has long been conducted between us, palynologists.
Of course human factors can influence the deposition of charcoals particles. But even in karst deposits, without the presence of man, the fire signal is not comparable. The most important is not to compare sediments with different taphonomic processes. In all cases, taphonomical factors must be taken into account.
Maybe you kwon the important studies on taphonomic processes modifying the charcoal assemblages in different contexts (Braadbaart et al. ; (2009), in Jour. Archaeol. Sci. ; Scott & Damblon or Théry-Parisot et al. in the special issue of Palaeogeogr., Palaeoclimat, Palaeoecol. (vol. 291, 2010) ; Marquer et al. (2012) in Jour. Archaeol. Sci.).