(co-author of the OCaml memory model paper here)

The details of the 'LDRF' (local data race freedom) property are described in detail here: https://anil.recoil.org/papers/2018-pldi-memorymodel.pdf

The performance numbers are in the paper abstract: "our evaluation demonstrates that it is possible to balance a comprehensible memory model with a reasonable (no overhead on x86, ~0.6% on ARM) sequential performance trade-off in a mainstream programming language". It's a little higher on PowerPC but still very usable, and RISC-V overheads should be roughly comparable to ARM.