One approach I'm finding useful is to deal with discrete projects/responsibilities/topics in regular time blocks, and within each, use a limited part of the time to "adapt to changing circumstances": emails, news, changes in legal rules (I'm a lawyer), etc. That keeps my focus on building/creating, not just reacting.Some changes must be taken into account. Some resolve themselves over time. Others don't matter.
It's also helpful to make a list of "all" possible inputs/tasks/focuses of attention — then, out of every 5, select 1. And out of those, select one of every five. That gives you (roughly) the top 4% of important things.
I also read something on HN today that resonated [1]: "you can only ever actually be doing one thing" so pick anything that "would be worthwhile to do right now, without any expectation that you know what might be 'best'", do it, and repeat.
FOMO is a b####, but virtually everything is unimportant. And with respect to important things, consider the unimaginable vastness of the universe and how trivial even the most important things are in comparison.
[1]: previous discussion: https://news.ycombinator.com/item?id=36253882