Regarding the analogy between mining speed vs. size of quarry, the number of steps shouldn’t be understood as the size of the quarry itself, but rather how much of the quarry you plan on using. If you have a quarry with 100,000 tons of resource, but your machine can only mine 1,000 tons within the timeframe you have access to, then the effective size of your quarry is only 1,000 tons, even though it has a size of 100,000. When increasing your speed of mining, it’s much easier to increase your effective size (step count) when it’s below the actual size of the quarry (dataset size), than it is to get up and find an entirely new quarry when you are already using the entire quarry (getting a new dataset).
Regarding the analogy between mining speed vs. size of quarry, the number of steps shouldn’t be understood as the size of the quarry itself, but rather how much of the quarry you plan on using. If you have a quarry with 100,000 tons of resource, but your machine can only mine 1,000 tons within the timeframe you have access to, then the effective size of your quarry is only 1,000 tons, even though it has a size of 100,000. When increasing your speed of mining, it’s much easier to increase your effective size (step count) when it’s below the actual size of the quarry (dataset size), than it is to get up and find an entirely new quarry when you are already using the entire quarry (getting a new dataset).