Native vectors types: GCC vs NEON

Is there an advantage to explicitly using the neon_vector_type over __vector_size__ when targeting ARM CPUs?
Because:
* according to the documentation:
http://clang.llvm.org/docs/LanguageExtensions.html#vectors-and-extended-vectors
neon_vector_type does not support subscripting while the GCC/__vector_size__ style vectors do
* using __vector_size__ everywhere reduces the amount of platform specific code...