Lines of code has very little meaning for me as a measure of "code quality". In any system, there will always be a minimum level of "complexity" that's necessary to properly encode and solve the software problem at hand. Reaching that theoretical minimum level of complexity a matter of expressing the logic in pure lambda calculus. I have no desire to read lambda calculus encoded real world problems.
My personal measure of "code quality" is all about readability that allows me to grok the intent of the logic such that I can make further updates to the logic in a safe manner. I want to the ability to grok bugs in the logic and otherwise tinker with the logic to improve the software towards the users' context.