Has anyone used multithreading/event-based operations for YouTube API?
My lab has a few of different account API keys (~30) and that extends our daily cap for YouTube API. Is it worth it to make API requests from different threads (in Python)? If it's worth it, I'm considering upping our number of API keys to speed up the crawling process. Since this is more network bound and I have enough CPU power, I'm also thinking of using event-based/async network op like Twisted.
I'd prefer not using distributed crawling since other people have their own tasks.
Edit: I've moved away from using YouTube API to using Selenium to get around daily quota.