The best, realistic POD I can think of is
an early invasion of Greece in 1940. It pretty much butterflies away the Blitz, and sheer momentum make make Crete a little less horrible for the Germans. Yugoslavia probably never has a coup as a result, so all of this clears the way for a slightly stronger German attack
somewhere around June 7th-10th, 1941.
The attack may achieve an identical element of surprise, because Britain is in the war, and mass encirclements in AGS become not just possible, but probable. After this point, Germany's best chance (presuming butterflies that do not lead to much larger than OTL German advances, or negative butterflies that the Germans advance farther but get encircled worse) is to prepare for a two year war, making Kharkov, Rostov, Sevestapol, Smolensk, and Leningrad first year objectives. They probably can't get Leningrad, but after this point, the Germans pursue an attritional strategy.
It can possibly be sold to Hitler as the historic end-times battle for humanity against mongrel hordes or some crap. Now, Hitler not looking for total victory on the cheap is ASB in of itself, but the OP seems to lend to this possibility.
If this occurs, the Germans find themselves without the heavy equipment losses of 1941, more aircraft, and 200,000 to 300,000 more experienced men, which is nothing to sneeze at. With Sevestapol in German hands early and Rostov capture and actually prepared for defense, a strike against Baku is likely in 1942, which really does not help the Germans. However, without the heavy 1941 losses, it makes the Germans all the more intimidating. Plus, much of their seige equipment probably gets tried out against Leningrad instead of Sevestapol in 1942, the result being that the Russians never break through ans establish a corridor to supply the city. This makes the fall of the city at least possible, or at the very least, things all the more horrific for the Russians.
Being that these "small victories" still do not definitely win Germany the war, it is easy to see why Hitler went all in. It's all in or nothing. Germany didn't want to gamble that they could win a war of annihilation if they had to. It is only when they found themselves in a war of annihilation by mistake that evaluating how much Germans were at killing Russians and kill ratios becomes relevant.