optimized convenance methods for conversion between SIMDe public types of the same size

[mr-c]

For example:

simde__m128_to_simde_float32x4:

• Not to be confused with simde__m128_to_neon_f32 which is only availabe when SIMDE_ARM_NEON_A32V7_NATIVE is defined.

See sfztools/sfizz#1163 for an example. In SIMDe 0.7.2 and earlier this wasn't necessary, but due to the way the public & private versions of the SIMDe types work, it would be nice to have these converters defined already (with optimizations, when available)

[mr-c]

Topics of discussion: are the optimized versions actually better than a memcopy?