At a minimum the question must reduce to physics. That is, when striped of context it must be answerable in terms of physics.
That implies that there are a lot of engineering questions which probably should not be on physics.se. Some examples of these would be
- Questions about regulatory issues. Legality and limits of various parameters.
- Questions about cost or cost/other-metric issues.
- Questions about engineering tradeoffs, as these necessarily involve decisions beyond physics.
That does leave room for some engineering/design questions. Strength to weight ratios, buckling limits, maximum deflection and so on. However, such questions have generally received little attention in the past. Most of our users have had limited exposure to the complexities of real-world analysis in favor of deeper understanding of microscopic physics.
That is to say most of us are physicists rather than engineers and though we share a lot of common preparation there is a big difference. Heck, there is a big difference even between the real-world preparation of you garden variety experimenter and a bog-standard theorist and the engineers and machinists I work with are always careful--when they first meet me to--to probe how much I know about the real world because they've met enough physicist who are simply dumb about real-world stuff to be wary.