In my software we deal with complexities in OSI layers 4-8, in multiple languages, with multiple protocols. As long as I've been working in core infrastructure software it's always been this way, and I think that's why I stay in the domain. The challenges are always new and this space sees a lot of constant iteration, but it's usually iteration down in the stack. We deal in systems technology all the way up to HTTP. There's a big variance to the nature of work you'll take on is my point. Additionally, this team seems to benefit from fundamental knowledge in math, systems, and software engineering.

When I worked strictly on web backends the most amount of Layer 4 I ever touched was a load balancer. I dealt in a few protocols but they were all fairly predictable. The system was not really something that we touched, improved, or even really relied on. It was more a consequence of development, so patch solutions were okay. I don't see the kind of fundamental knowledge on my team being required when I filled these kinds of positions, but maybe it could.

The thing I have trouble reconciling is whether I judge web backend development more harshly (and thus think, this is not the same work) purely because it wasn't interesting to me, or because my company doesn't mandate that kind of attention to detail (or tries to provide broad solutions and discourages individual problem solving), etc... It's hard for me to tell on an individual level.