I think it's virtually impossible to identify Stiphodon species for certain by looking at photos. As Seriously Fish says "The Gobiidae is the most speciose vertebrate family and notoriously problematic in terms of identifying fishes down to species level.". The identification of species is basically down to anatomical features such as number of rays in various fins. Consider Seriously Fish description of one species "It can be told apart from congeners by the following combination of characters: first dorsal-fin with 6 spines and 1 soft ray; second dorsal-fin with 9 soft rays; male with pointed, but not filamentous, first dorsal-fin with spines 4 and 5 elongate; posterior tip of first dorsal-fin normally extending to base of soft rays 2-5 in second dorsal-fin; pectoral-fin rays 15-17 (mode 16); premaxillary teeth 32-40; dentary with 32-45 horizontal teeth; female with 2 dark lateral stripes with a relatively large space between, lower stripe comprising 8-11 regular blotches, upper stripe thin and pale; all fins of female usually with faint markings; non-nuptial males with a dark lateral stripe below lateral midline, absent in nuptial males; nuptial males with blackish or sometimes orange body, metallic turquoise laterally on head; metallic turquoise spot on upper pectoral-fin base in life; pectoral-fin rays in male often having 1-6 somewhat obscure black spots."
Of Stiphodon they also say "Over 30 (species) have been described to date with at least a similar number awaiting description"
Also consider the distribution and lifestyle of these fish, most are found in very short streams that run down to the sea, the larvae develop in the sea and swim upstream later. Specimens from nearby streams may look different, but actually be varieties of the same, or they may look the same on the outside but actually be different species. Some of these islands have a very large number of streams, there could potentially be many Stiphodon species on even a small island.
As a professional invertebrate surveyor, I meet a similar problem regularly, people insist that the insect they have seen must be a particular species because "it looks the same as pictures on the internet (or in a book)". They rarely believe me when I say that there might be many species which apparently look exactly the same, but once under a microscope (or even dissected) the differences become clear. People joke about us "counting the hairs on a fly's legs", that's not quite true, we don't count the hairs, only the bristles
A similar situation applies to these less well known groups of fish, looking at pictures may not give you the correct answer.