Improve Python bindings and Windows&Apple Silicon support

[ThePituLegend]

This PR originally aims to improve the initially proposed Python bindings. They were functional but not easy for the final user to use. Thus, we streamlined the Python experience as much as possible, as this feature has been deemed useful for the community. There are still some rough edges, but it is far better.

To that end, we worked on the following features:

• Python module is now compilable and installable from source code with a simple pip install .
• We provide precompiled Wheels and SDist, which are hosted at PyPI. Thus, end users can install QuickEd for Python with a simple pip install quicked.
  If there is sufficient interest, we will consider working on preparing Conda-forge/Bioconda packages.

To handle this new setup (mainly for precompiled Wheels), we implemented a dynamic dispatcher in the Python module that chooses between the SIMD or scalar compilations depending on host availability.
This process indirectly forced us to add better support for Windows and Apple Silicon machines, as we should support many configurations for the Python package to work on as many machines as possible.
Now, the CMake default config considers the following targets (OS/Arch/Compiler combo):

• Linux (Ubuntu) x86_64, GCC & Clang
• MacOS x86_64, GCC & Clang
• (NEW) MacOS ARM (Apple Silicon), GCC & Clang
• (NEW) Windows x86_64, GCC, Clang & MSVC compiler

Unofficially, Android ARM also works, although there is no precompiled package for this target.

As more platforms are now supported, we updated the release Action to upload precompiled binaries for all supported platforms automatically. We also added a release Action to build and upload the Python package to PyPI automatically.