The same results can be derived from different instructions. By simplifying; you can perform multiplication by repeated addition, or combinations of both, or even more clever tricks which would hardly make the original intent of simply making a multiplication obvious.
Running the code in a simulator would give you the final output, but would do very little to explain how the code itself is working, which is what you usually are after.
Okay, so this is simply an advanced form of obfuscation.
An hypothetical uberman can still reverse engineer it. I don't expect anyone to, but the article seems to claim that there is some mathematical impossibility to do so.
Running the code in a simulator would give you the final output, but would do very little to explain how the code itself is working, which is what you usually are after.