I think it's pretty much impossible to tell statically whether you're going to be drowning in overhead spinning up a thread for every pixel on the screen, or le

I think it's pretty much impossible to tell statically whether you're going to be drowning in overhead spinning up a thread for every pixel on the screen, or leaving tons of parallelism on the table. Maybe with PGO it can be made to work. Even if it could be automatic, parallelism can be so important, and usually requires explicit design to enable, you probably don't want to delegate it.