I blind-tasted wine competitively at an international level and it is pretty commonplace to identify region/sub-region/year/grape across 10 wines blind with ~80% accuracy. In fact, many tastings go further than this and you identify chateau/year from 10 wines from the same region - i.e. the Oxbridge teams mentioned there also do a blind tasting competition specifically on left bank Bordeaux. It's relatively straightforward to train yourself to do this. Note that the physical properties of the wine play heavily into this too (colour, reflection in light, viscosity are proxies for alcohol, age, grape, treatment). Additionally, having the requisite knowledge of "this wine would never be served this old, this wasn't a good year in this region because of x weather, this region only grows x,y,z" means you can deduce a lot from quite little information.
I do think people overfit on the colour piece though; you could definitely find some reds and whites which, if tasted without seeing the colour and at the same temperature, would be challenging to tell apart, more-so than differentiating between country/grape/region. I'm not sure what this really proves, apart from being counterintuitive.
That said, there is a lot of bullshit in market of people who buy and sell expensive wine (more so than the professional tasters and vineyards, who are usually genuine enthusiasts). Lots of cheaper wines may be more enjoyable, and wine flavour profiles are often full of weird signalling about what constitutes quality, i.e. "tasting note x is considered desirable/refined/distinguished, ergo it tastes better", which is nonsense and snobbery. Without any bias you might think expensive and old bordeaux tastes like an old boot, but if you've been trained that those specific flavour profiles are distinguished and desirable ('hints of leather and earth').