1. Unless refactoring existing code makes new code easier to write, or fixes existing but broken functionality. The design of least resistance may not be obvious the first time a section of code is written, and patching around the design is bound to result in less rapid progress on features than just refactoring immediately to save time and pain later. Not all time spent writing new code is time well spent, unless there is a line-of-code performance metric that needs to be gamed in order to
appear productive.
2. Experience can be a bad thing, depending on how one relies on it. It depends on whether tone relies on rote memorization and verbatim application of things learned by experience, or whether one spends time analyzing what one has learned to understand when what was learned is applicable. It is quite possible that BigCanOfTuna's complaint is legit, that his boss writes code that requires a lot of additional work and attention. Or it could be he is refactoring because of trivial stylistic things, or just to refactor. I don't know, but I'd like to give him the benefit of the doubt until he provides reason for me to believe he's just wasting time and trying to get his boss's hackles up.
3. And ultimately, you sometimes have to ask whether loyalty to your boss or the success of the company is more important, because these are not always mutually achievable goals. I may be reading too much into this, but you seem to be advocating a monochromatic, obey or leave, line of thinking.
At BigCo, it wouldn't be disastrous if you followed bad direction from a boss. You would be trading an insignificant amount of financial damage for political stability and security. I would still have a hard time doing this and would probably be quite demoralized at an employer that demanded it, but I can at least understand others aligning their priorities this way.
At a startup, unwavering loyalty to one's boss can potentially do much more damage, and deciding between financial success and political security isn't so easy, when forced to pick one or the other. The decision can
become even harder when one has bought into the vision of the company, and doesn't view what they do as just a job to pay the bills; they might not be prepared to leave just to avoid a struggle. This may sound hopelessly idealistic; so be it.
Ultimately I don't know what BigCanOfTuna's workplace and boss are like, what his priorities and goals are, and whether his complaints are legit. I do think he deserves more than a cursory dismissal as just another whiny kid who has to get his way.