I think there's a whole lot of truth in this, but I'd also like to try to provide a complementary position, for the balance.
There is a kind of depth can come from focus, but this is a narrow kind of depth, one that can't question its own premises. It's directly analogous to the depth of a depth-first search, and has the same upsides and downsides. It's really handy when you know your target, and when you think your heuristics for having chosen that target are pretty good.
Meanwhile, I think you can look at this kind of broad-spectrum curiosity as a kind of breadth-first search, again, with the same strengths and weaknesses. It definitely takes longer, and it definitely can end up amounting to not much more than wallowing. But it can also serve very well when you're not sure what the target is, or if there even is one.
Furthermore, it can give access to more possible models, more analogies, more metaphor, more understanding-by-comparison, which in turn grants a depth of its own kind, especially in conjunction with true narrow depth of exploration. Since all models are wrong, and each model explores different facets of an idea, having more models gives you a broader set of tools for taking a given concept and applying it to specific areas. They may not be useful in the given moment, but they may lead to unpredictable and insightful connections later on.
But of course, to your point, that all comes at the risk of never getting far enough along to apply anything in the first place. But I do believe there is a place for each. Feynman spinning plates in the cafeteria and all that.
There is a kind of depth can come from focus, but this is a narrow kind of depth, one that can't question its own premises. It's directly analogous to the depth of a depth-first search, and has the same upsides and downsides. It's really handy when you know your target, and when you think your heuristics for having chosen that target are pretty good.
Meanwhile, I think you can look at this kind of broad-spectrum curiosity as a kind of breadth-first search, again, with the same strengths and weaknesses. It definitely takes longer, and it definitely can end up amounting to not much more than wallowing. But it can also serve very well when you're not sure what the target is, or if there even is one.
Furthermore, it can give access to more possible models, more analogies, more metaphor, more understanding-by-comparison, which in turn grants a depth of its own kind, especially in conjunction with true narrow depth of exploration. Since all models are wrong, and each model explores different facets of an idea, having more models gives you a broader set of tools for taking a given concept and applying it to specific areas. They may not be useful in the given moment, but they may lead to unpredictable and insightful connections later on.
But of course, to your point, that all comes at the risk of never getting far enough along to apply anything in the first place. But I do believe there is a place for each. Feynman spinning plates in the cafeteria and all that.