This course was absolutely worth it. I always had this idea of online certificates being diploma churning entities that will pass everyone, b/c they are looking to get paid. This certificate was the antithesis of the aforementioned. Get ready to put the time and effort if you want to do well.
As an undergrad I took two semesters of C++, and I thought that taking this course would not be worth it; that I should go straight to the advanced certificate. I was wrong - by the fourth level we had already covered everything that I had gone over as an undergrad (and in much greater detail). Crucial concepts like inheritance, polymorphism, generic programming, and the usage of both the STL and Boost libraries were meagerly, if at all, mentioned in my undergrad courses. You will also learn an array of implicit ways of how to make programs that are nimble and efficient: inline functions; macros; colon notation; static vs dynamic binding...you name it.
Then you also have to consider the invaluable asset of having expert TAs providing feedback in a matter of hours after submission of both homework assignments and questions; I never felt I was stranded in an island with no one to consult if I had an inquiry.
But quite frankly, the most empowering part of the certificate was the final project. As a Fin Math major, I have come across countless instances during my courses where I would fully understand the math and finance behind the material taught, but when it was time to apply things programming-wise I would have a hard time wrapping my head around things like how, for example, can a stochastic process modelling the behavior of a stock be replicated in a computer? I understand the idea of sampling from a given probability distribution, but what does that mean or look like when it is time to program it? The expected value of a random variable, of a payoff function, in a computer? How?