Usually #1 and #2 are not in conflict. The whole role of a manager is to translate company goals into engaging team projects.

The point I think the author makes is that it is not the desires of the team that set up priorities at this point, and that it is contrary to what you are expected to do at the dev level (i.e. advancing your team projects first)

The thing is, if a manager does not set goals for it's team, it will find some by itself (refactor that old code, clean the doc, improve that funky demo, test out the latest shiny equipment) that are not necessarily the best choice for the company.

My experience has been that in that situation, very often, the team is relieved when it receives new instructions. It is not a "team or company" thing It is just an ordering of priority goals that makes sense to everybody.