Clean code minimizes these measurements.
To minimize variable count I can use ternary expressions. To minimize indentation level I think carefully about early returns or extracting blocks of code to functions that do nothing if condition is true (early return).
To minimize grokking time, I revisit old code I've written and measure the time it takes me to understand it. It turns out it correlates quite a lot with line, variable, and token count, indentation level :D
Clean code as commonly defined is far away from any measurable metrics. That's why one might say that it does not exist.