A few observations, but it is up to you to assess the relevance.
Recruiters are simple beings that are easy to predict: they will say anything to get you to take a job that's on their books and for which they might get extra income. That doesn't mean they are incorrect.
In software engineering there is a fashionable strong movement to "devops", i.e. don't have separate design, test, operations teams in separate silos. While there is a degree of transient fashionability in that, the underlying concept is good: remove silos and make a team out of all the individuals with the relevant skills.
Development is a "red queen's race": you have to run as fast as you can just to stay still. If you are out of development, then returning risks your starting on the bottom rung.
When I was at your stage (i.e. first job for 3.5 years), I knew I didn't know whether I wanted to stay in R&D forever, or whether I woudl like to move into technical sales/marketing, or project management, or up the greasy pole. I found myself a job which involved elements of all of those, at a contract R&D house (Cambridge Consultants Ltd). That was a good decision since I was able to sample everything from analogue to semi-custom to hard real-time software, plus writing proposals and Ts&Cs, plus leading projects, plus client interaction.