I don’t know how they fared in the old days, but in more recent years, when the Seminoles encountered the Hurricanes, they usually hung right in there until the very end, when the 'Noles would veer wide-right or wide-left…
Sorry! Seriously, though, unusual cloud patterns in advance can be somewhat of a tipoff, even though it’s true that in the past, people were often taken largely by surprise. I think it partly comes down to exactly what time period people consider “early”. Before Hurricane Andrew hit South Florida, you could tell that a big storm was approaching some 12-18 hours or so before it made its landfall: we were beginning to get some sustained winds from the east, the sky was darkening, and, perhaps most tellingly, the birds were all heading west as of that time. If that’s all the warning time you have, then any metropolitan area (such as Miami in 1926 during Camille) would be caught mostly off-guard, but perhaps that was enough time for, say, the Miccosukee to strike camp and canoe to a marginally safer area.
In retrospect, I wish I’d paid a bit more attention to the environmental aspects of the storm’s advance (just out of scientific interest) and had made it a point to look for certain early indicators, but I was just too busy helping my parents secure the house and office. But I do have a distinct memory of noticing the birds’ migration that afternoon, as they were crying (calling? you know what I mean) to each other and flying very fast, very purposefully. They’d probably been flying west all day (and possibly the previous day), but I only noticed it then. Ditto for the cloud patterns; there might’ve been good indicators even earlier, but I wasn’t paying enough attention or taking notes.
One more thing about the birds evacuating in advance of a hurricane – the last bird in the Everglades to bug out is the ibis, which is why that’s the mascot of the U. of Miami. (Whether that’s indicative of bravery or stupidity, though, is a matter of debate. :wally )