We can now use for e.g. vec4<float> instead of float4, this is useful for templated code. The existing solution was to reach out to the math::detail::TVec<> class.