It's not technically hard, no. However, there's an overhead in running a headless browser, of memory, CPU, time, and bandwidth, so it might cost substantially more.

You don't need a browser for everything, but working out which pages need a headless browser and which you can get away with scraping sounds like an interesting problem.