When a line is configured for G.INP, some issues are still best dealt with by the FEC+interleaving process, so this can still be configured. I've certainly seen a lot of cases where FEC+interleaving is part of the G.INP configuration.
However, I've found that the FEC overhead is much reduced (around 5% instead of 20%+), and the latency is much reduced (around 0.2ms instead of 8ms).
Why? G.INP is well suited to handling SHINE, but would be overloaded by REIN. The new set of settings for G.INP allow the two to be handled separately.