Is there any possible justification for an instruction with two opcodes?

An instruction only has one opcode.

There may be two opcode bytes, or there may be different fields in the assembly code of the instruction, but the bit pattern of an instruction will always generate the same results.

There are some results that can be generated with two different opcodes. SUB A and XRA A, for instance, both clear the accumulator, but they are two different opcodes for two different instructions.