I have a hard time classifying this as a "design problem" when it has zero practical implications. I've never encountered a problem that can be traced back to this, nor ever helped anyone else with a problem that can be tracked back to this. If it is a "design problem" it is of the weakest possible class. I can't imagine what would "fix" it that wouldn't be a larger design problem of its own.

The whole "confused about nil interface vs. interface containing nil pointer of some type", while I would contend has a root cause of improperly importing the idea from C++ that nil is automatically invalid when in fact it is perfectly valid in Go to call methods on nil and thus confusing "nil" with "invalid", is at least a cause of real bugs and real misunderstandings.