Hmm...in order for me to really explain to you why robust PCA is needed. I will have to explain to you what PCA is. In order for that, I would have to explain why? High dimensional spaces with lower dimensional information, understand what dimensions are, what a basis is, why variance is a brittle measure, why outliers break it, so we get to robust PCA.

Or I could tell you that I find better ways of extracting information from data.

Your pick. The latter just sounds like shit. But just because you have a degree in computer science doesn't necessarily mean you will have the tools to understand that.