Confusion can arise about the appropriate analytic method when camera arrays or other forms of non-physical capture constitute the re-capture samples. Here it is important to keep in mind the nature of the marking method, with respect to the putatively unmarked animals:
When unmarked animals remain unmarked-- such as the case where unmarked animals can never be individually distinguished-- marked-resighting methods are appropriate
When all animals at least potentially are "markable" -- whether physically handled or not-- ordinary CMR (or spatial CMR variants) may be appropriate. Such cases naturally arise when animals are individually distinguishable via the sampling method, e.g., by antler charactertics, stripes, scars, or genetics markers.
In some cases a hybrid situation exists, in which some animals are unambiguously identifiable; other can be classified as "marked" but not identified to individual; and still others must be considered as unmarked. Some of the mark-resighting methods handle variants of this, but research is ongoing as to the best ways to deal with this sort of data.
In any case, success of these methods depends on getting decent number of resighting of marked animals. Studies in which few marked animals are ever seen again but large number of unmarked animals are seen in resighting samples typically produce very poor estimates. This makes sense because at the end of the day we are using a very small sample of M to get at the number of unmarked animals out there, and that is going to result in very high variances and sometimes nonsensical estimates.