Abstract algebra is more relevant to the general practice of cryptography engineering than semiconductor engineering is to the general practice of writing software.
I'm sure it is but it also has negligible application to quotidian grunt-programming which is regrettably what 99.9% of people on HN do. Learning a skill that seems never to get used seems completely pointless to me. That's a critique of industry, not of linear algebra by the way.
I agree that abstract algebra has only marginal important to the general practice of programming. I only dispute that it's marginal for cryptography engineering.