The visuals can look quite good at any given moment. Buildings are rendered superrealistically and look equally sharp at a distance and zoomed in, where you might get close enough to see details like a variety of foods at different street vendors. The cities have a realistic scale, which makes the Soviet architecture of institutional buildings look all the more imposing. In contrast to all that detail, people have a caricatured look. But the way they move is what's really off-putting. At the beginning of a phase with an important conversation, the camera will lock on an odd pathfinding dance, with characters going out of their way on indirect paths before finally coming together for a friendly salute. Such events aren't hand-scripted--you pick the specific location--and it's obvious. Other story events display similarly awkward pacing, with drawn-out set-up shots of 20 seconds or more just showing cars rolling up and a few peripheral people getting out. It wouldn't be quite so tedious except for the fact that the camera is locked for the least entertaining and least interactive parts of these sequences.
