I think you have a point, but it's not a terribly surprising one to most people here. This is pretty much the same thing people had already realized with Node.js, after all. I'll even go as far as to say the use of uvloop is fair, it's a reasonably generic piece of software that applies to a wide range of problems. It fits the Node.js comparison well, too.
But personally, I don't see how the usage of hiredis is so much different from literally taking a part of Redis itself and embedding it in Python. As a thought experiment, if I did take Redis's parsing code and embed it in Python, would it still be fair?
I'm not really convinced that a pure Python implementation of the Redis Serialization Protocol would actually be meaningfully slower than the Python bindings to hiredis, so this whole thing may be a bit moot.
I think your project is cool. Perhaps some changes to the README to explain exactly what your benchmarks are meant to show about Python itself would help the criticism.