In Vol2 of “Paul Keres’ best games” by Varnusz, a lot of games are won by smoother, better development in the opening than his opponent. I’d say most of them involve marvelous sacs, but I am going to show one that doesn’t, that illustrates the idea of better development more simply.
By move 15, Black is down two pawns and probably already has a winning advantage based on superior development. By move 20, White has played the final blunder by playing 20.Nxe4?? (allowing Black to capture a developed piece) instead of 20.Nf1 (allowing for further piece development by freeing up d2 for another piece).
The thing to note here is the question “How many Class A players would think they have a winning position here as Black while two pawns down?” There were many other possible continuations during this game where this same sort of advantage had to be measured as well, development vs. material. It seems if the queen is running around to grab pawns while Black develops, then -+ or =+ is frequently the result. How many class players can make this type of assessment? Very few, and the few that can probably have weak endgame skills or they wouldn’t still be Class players.