In my humble opinion the behavior you expect is more like an engineer than a scientist. The latter might be involved in some innovative research project or the exploration of yet unknown possibilities of some new technology. The former works along the lines of standards and best practices to build something that is reliable, maintainable and has predictable behavior in nearly every environment.

The science behind databases has mostly been developed by mathematics. While I think it is necessary to understand the theory behind the tools you are working with on a daily basis, one does not need to be a scientist to maintain a production environment. Both the artist and the scientist should work in a development environment; even the 'art' of performance tuning should not be conducted within a production environment. As usual, it depends ... in smaller organizations resources may be scarce and the artist, scientist and engineer may be the same person.

Working as an engineer pays of when your work has to last for a long time and people are willing to pay for all the extra time involved in this approach. In every day life that might not be the case: things must be adapted fast to new technology and new demands of clients at the lowest possible costs. That's the hardest task today in guarding your production environment like an engineer ... convincing your boss.