A larger context window means a model can read more text in a single conversation. It doesn't mean that text stays current, that it's connected to your other tools, or that anyone else benefits from the work of assembling it. Those are three separate problems, and a bigger window only ever solves the first.
For a genuine one-off — analyzing a document you'll never look at again, or a dataset that has nothing to do with your ongoing business metrics — pasting it into a long context window is faster than setting up a sync for something you won't reuse. The trade-off changes as soon as it's a question you (or your team) will ask again: a synced connection does the export-and-format work once, instead of every time.