One should keep in mind that the map is 25 years old, meaning it doesn't really show the post-soviet development of Eastern Europe or the population growth of Asia and Africa. But I don't think that the distribution of "empty" (less than 10 people per square km) has changed much, so for this purpose the map is perfectly adequate.

There are areas of grey (less than 2 people/sqkm) around the coasts everywhere. A processing artifact, or some indication of low water marks? I don't think the tidal area can ever be 20 km wide, but it's very sloppy if it's just bad drawing. (There's already a cyan outline for the coast! Why add grey areas too?)

Then I'm certain there are areas in Wicklow, Ireland with less than 2 residents/sqkm [0]. But the whole area is shown not one but 4 levels more populated, at 100+ residents/sqkm, comparable to Las Vegas and ahead of Buenos Aires which is shown at 41-100 residents/sqkm.

The resolution of the map appears to be at the 10km level, but I think this is misleading and in most of the world it's actually based on data at the country level or one level below.

One of the most surprising things to me about the map before I looked closely was the lack of wilderness in Turkey. If it's that dense all the way through, it should be a massive power. But aren't there mountains and deserts? Probably just the map being lazy again.

[0] https://maps.app.goo.gl/cktt79N4FDLegF7a6 - not the most egregiously incorrect measurement on the map, but one I know personally