Optimize performance when gathering content pages

[o1y]

Reimplement #154. Content gathering could take a long time for large Statamic sites. This commit makes these calls concurrent if spatie/fork is available.

In this implementation, URLs will no longer be cached in SSG because Statamic 5 already has an optimization for it.

[jasonvarga]

I'm assuming you're using the Cache there because you had trouble returning big collections from the fork tasks?

[jasonvarga]

I added usleep(500) into Entry::uri() to simulate a slowdown and I don't notice any difference between the master branch and this one, when using workers. 🤔

[o1y]

I added usleep(500) into Entry::uri() to simulate a slowdown and I don't notice any difference between the master branch and this one, when using workers. 🤔

Oh, haha, you're right, there's no difference anymore - sorry! That was actually one of the main reasons I decided to close the PR since V5 has already addressed most of the slowdown issues when gathering pages, so this PR is no longer necessary. 🎡