Comment by canofbars

The point is not to test everything ever published, its that when you want to do X, you look for papers on X understanding that they are likely flawed but better than starting from scratch.