I would argue that your central point is intrinsically flawed. Your claim is that starting at a high alcohol percentage, then gradually decreasing alcohol content, gets you intoxicated faster. @nico is correct that you will get drunk faster this way (more alcohol sooner rather than later, so this is trivial). If we allow only one variable to change, the alcohol content (and therefore the water content), then the first drink is the most potent, and each drink afterwards is getting progressively more dilute, then this implies you are able to rehydrate better the more you drink. This argues exactly opposite to your claim.
Since you offered up the example of vodka to wine, this introduces two new important aspects: fermentation impurities and sugar content. Vodka is distilled and relatively pure, so it offers little in addition to alcohol and therefore easy to process by the liver. Scotchs and whiskeys are low sugar, higher impurities, and so would be a bit more troubling to filter and process. As you move down to wine, impurities rise and sugar content is much higher. This added sugar load dehydrates faster as well as causes a cascade of hormonal changes that are more immediately troubling than the alcohol. Beer would generally be the extreme end of the impurity and sugar continuum.
Since you bring up the USSR and the KGB, I imagine that if they did stick to this regimen of interrogation, it could be more effective by power of suggestion. The victim believes they are more intoxicated than they are, and perhaps more likely to offer up their secrets.