Seems it is actually on the other side of the map.

It's not cut off. The problem is it's not being rendered on each repeat of the world. As a different example

The polygons should be seen 6 times in the screenshot. The static map gets this right.

There is apparently a Leaflet option to fix this problem. It is called "worldCopyJump"

https://leafletjs.com/reference-1.3.2.html

Hmm, that doesn't really solve it though. Known issue:
https://stackoverflow.com/a/40544502

This is also relevant to T340084 Wrong rendering of <mapframe> on transition past 180 longitude due to inconsistency in longitude handling

Please note: This is not quite a duplicate of T340084 as the issue is not purely upstream and the upstream issue is now marked as closed. T340084 mentioned that handling was different if the calling wraparounds were centred West or East of 180E/180W and the imaging thumbnails were rendered with inconsistent behaviour. The issue will also if worked through to ensure Wikipedia consistency in rendering require a check against the reimplementation of the Graph Module when Template:OSM Location map gets back its useful functionality. My assumption is that the Wikipedia wraparounds could be programmed to handle the upstream issue given the debate before the upstream issue was closed. Cheers

I'm not 100% sure it's the same, but I suspect it is. The differences in rendering are explained by different engines being used for pageviews and previews.

I have noted due to a user complaint (see [https://en.wikipedia.org/wiki/Template_talk:OSM_Location_map Template talk:OSM Location map] ) that this issue is assigned low priority but perhaps is higher priority for editors around the Pacific Rim. As said part of issue is different backdoor engines but still likely inconsistent wrap around code.