The issue lies with how you emulate the loop indices. You're conflating the condition evaluation count with the loop body execution count. The first loop executes 5 times, and not the 6 you have. The second loop also executes five times only, and not 6. The condition is evaluated 6 times, but not the loop body. Since you're trying to emulate what's going on, you should make sure every loop apart from the last one only counts the number of loop body executions - the innermost loop should be the only one that executes +1 more time. Correct these and you'll correctly emulate the 125.

The last loop's condition will be checked 5 times since it's inclusive. It's enclosed within a loop that executes (from here on we consider loop body executions NOT condition executions) 5 times. THAT loop is enclosed in a loop that executes 5 times. The result is the multiplication of all of those: 5 * 5 * 5 = 125.

IMO you should write whatever matches the idea you have. With time and practice you pick up shorter ways to describe what you have in mind.

Python list comprehension tends to be faster than the equivalent for-loop, but that is probably not the reason why people are choosing to use it.

You are re-implementing some common patterns, min and map. When you start recognizing them, you might start using the already existing implementations instead.

List comprehension combines the functionality of map and filter, but the filter part isn't needed here and the mapping function already exists - that makes map the minimal already existing thing that fits the problem.