Yep. For returns, the more important thing in the article, the "ret"[0] instruction behaves exactly identically to "br x30"[1], just with the hint that it's expected to match a "bl".
On x86 things are less pretty as call/ret push/pop the ip from the stack, with no non-matching-hinted versions, but having such would also just not be particularly useful as unpaired call/ret would result in the stack pointer continually drifting without manual fixup (at which point jmp is just clearly better).
On 32-bit x86, there is the trap of trying to use call/pop to get the absolute instruction pointer. It will work correctly, but it will mess up any call stack prediction and cause great performance penalties. Hence why compiler-generated EIP shims use call/mov/ret instead. (But this is not such a big issue in x86-64 with its RIP-relative addressing.)
