The benchmarks are open source and adopted from competitor libraries, standard tests that style libraries have used for a long time. Nothing deceptive.

The other benchmark posted tests more framework level stuff, but could be repurposed to show a more “typical app screen” with some effort.