Good question.
It's a good idea to understand programming if you want to trade systematically - even if you have no intention of actually building your trading system from scratch.
So learning a language for which there are extensive finance libraries to aid building a system and to learn the ropes is invaluable.
And you are right, Python is one of those languages - but it is not the only one.
Starting at Quantopian is a great idea, even if you have no intention of trading stocks. It will show you what the foundation of a trading system looks like. And with that knowledge you can incrementally augment your design and development skills to build what you like yourself.
And in any case, I understand you can now use your own data to integrate other products.
Try to keep your systems simple (whilst avoiding rookie mistakes). With luck you have a chance of succeeding. Above all, reign in your expectations. Consider your efforts a resounding success if you manage to retain your capital intact ove your first 5 year period.
Targetting 20% returns per month will only guarantee a blowout. Pretty quickly - if that's any consolation.
In my opinion, most people are too afraid of making mistakes, of somehow "looking bad". School and work promotes such fears every day: At school, pupils are expected to simply regurgitate what other people discovered through trial-and-error themselves, and to avoid attempting to create alternative answers themselves. At work, workers are expected to keep in line and bow their heads, as anything else is risky and rarely worth the potential trouble, especially for cross-cutting concerns.
How costly is the mistake? If it's endurable and don't affect you too much - both psychologically and financially, it's GOOD to recognize a mistake and correct it. In fact, being able to rapidly recognize and fix your mistakes is paramount, as I doubt anybody will be able to create a viable system without it. So be honest, ie. with yourself, otherwise one gets stuck in far-fetched fantasies of making millions or being able to make perfect designs/processes up-front, which is not realistic at all.