I'm a professional developer with a post-grad degree in Mathematics.
There's an ounce of truth in what he says, not the part about computer scientists or software engineers somehow being better than scientists, on the contrary, you're largely right because most programmers decry maths and claim it doesn't matter to them, but they're really just the dross of the industry. Maths is what separates someone reinventing the wheel by condemning themselves to produce CRUD applications for all eternity from someone who comes up with genuinely new and novel bits of software. Those with an understanding of maths are the ones who give us everything from the highest quality programming languages to Google search, and increasingly beautiful game engines to AI solutions like Siri.
But I digress, the point intended on making is that the fact is that those with computer science have been the ones helping push the largest gains in science in recent years, whether it's the type of data handling required at the LHC or entire subjects like bioinformatics. We've long passed a point where much scientific discovery can come from lone individual geniuses, and are entering an era where many problems are impossible even with merely teams of people. We're at a point where leveraging computing power is essential to much further scientific discovery, and for that you need computer scientists who both understand the science, and the machines needed to drive the discoveries in said science.
In this day and age I'd pity the scientist who looks down on computer scientists, because frankly in most areas of science it means they're building their own path towards irrelevance and failure. It's unlikely they'll ever achieve anything in most scientific fields if they're not willing to work with those who understand how to command the machines, or who do not learn themselves to command the machines and themselves become computer scientists in the process.