The real problem solved by BWX wasn't shifting and masking necessary in handling 8-bit values, as that was handled pretty transparently by compilers, it was how complicated it made dealing with hardware I/O which often expected 8bit or 16bit atomic writes - whereas Alpha could not do shorter i/o than halfword, i.e. 32 bits (All I/O on Alpha was MMIO).

This is visible in form of big split in Alpha hw support (drivers etc) on VMS and OSF/1 between "non-BWX" and "BWX" hardware.