Why do we drop constants and lower-order terms, so O(2n + 5) becomes O(n)?

Big-O captures the growth trend at large n, where the dominant term swamps everything else and constant multipliers do not change the shape of the curve.

For big n, 2n + 5 grows essentially like n, so we call it O(n). Constants and small terms matter for real wall-clock tuning, but for comparing how algorithms SCALE they are noise, so we drop them to focus on the order of growth.