consolidation of markers already occurs when observations are close enough to each other. so there is already some sort of threshold set for this in the current mapping interface.
probably the easiest way to accomplish this without changing the mapping interface too much is to simply allow users to select whether they want to see the grid markers, the pin + obscured circle markers, or the heatmap. right now, the grid markers show up in the Explore page map at zoom levels 2 to 10, and the pin markers take over from zoom level 11 and on. the grid markers and heatmap do provide an idea of observation density via opacity and color, respectively so if you could use those rather than the pin markers in some of the more granular zoom levels, then you could get an indication that a particular spot has a higher density of observations.
(if you want to see what the different observation map tile sets might look like at different zoom levels, you can look at https://jumear.github.io/stirfry/iNat_UTFgrid_based_density_map_for_Leaflet.html. this page shows some of my own custom density markers by default, but you can use the layers selector in the top-right corner of the map to see iNat’s standard grid markers, heatmap, etc…)
another thing that could probably be done relatively easily is when you click on a particular pin, you could have the pop-up show not just the latest observation’s details, but also additional text at the bottom that could be something like “+10 other observations”.
i don’t really advocate new “symbology” to indicate multiple points because at lower (less granular) zoom levels, sometimes most of the markers on the map might represent multiple observations. so the only time such symbology might be useful is at really granular zoom levels, i suspect.
there’s really not an elegant way to do this technically in the current mapping interface without totally changing how it works, i think. but even just conceptually, exploding points is fine when you have a limited number of points, but when you have the potential for hundreds or thousands or even more observations in a given spot, then exploding points might not be the best way to present such data, i think.