Unimportant to the business is not the same is unimportant for other purposes. There are a lot of unimportant projects that are interesting to an engineer.
Trying all the latest javascript framework is not important to the business, but but is your best person is free to try each fad, every once in a while he will be able to say "this one isn't just a fad we should use it". It is thus important for someone have time to do this.
Trying some new tool is not important, but it might long term result in better quality if the tool turns out useful.
Upgrading the CI system might not be important, but if you let it fall out of date you get into trouble.
Stepping back from the day to day schedule allows the best engineer to focus thinking up a better solution. There are many things that are done "because we have always done it that way", but in fact the pressures of a 8-bit CPUs and limited memory 40 years ago drove that decision and there is a better way that modern CPUs could do now if only people weren't blinded by how it has always been done. The proof of concept will get management excited enough to make this the number one project, but right now they can't imagine the better way is even possible. (and the best engineer will sometimes discover the algorithm isn't actually possible)
In short the unimportant projects the best engineers work on should be largely self directed. There is the following items that are a consideration so long as they don't take up too much time. (if they add up to 25% you need someone assigned to that problem)
The best engineer is also free when a customer discovers a show stopper bug. If it is easy he can get the fix done in a couple days without interrupting the main team.
The best engineer is free to tell marketing which ideas are possible and which is impossible. He can also help workout timelines. Thus ensuring that the next project is actually possible in the time given.
The best engineer is there when sales says "I could make this big sale if only we had feature..." - some of those features can be done in a few hours and make a lot of money. Others take years and lose money - but the salesman has someone to ask before making a promise that cannot be met (that doesn't mean the saleman will ask)
This is also a rotation - your second best developers will be playing "leapfrog" to the best and then your best developer gets rotated into a valuable assignment while the new best engineer gets to lead a project again.
If your company has any size to the engineering team you can find those engineers who are self directed who will enjoy the lest important project