At the low end of the scale I agree. But once you start scaling it's very difficult to run more than around 40 concurrent sessions with a browser-based scraper.

At the low end of the scale I agree. But once you start scaling it's very difficult to run more than around 40 concurrent sessions with a browser-based scraper. With a pure Nodejs scraper we can run over 1000 parallel sessions per CPU.