Is possible, with PERC H710, to replace a (working) member with another (newer) disk, without loosing redundancy during the whole rebuild phase?
Almost all software raid (ZFS, MDADM) does this. During a disks replace, the old disk is still part of the array and is working properly. In backgroup, raid will sync the new disks with the old one and only after a proper rebuild, the old disk is decommissioned.
Something like a RAID1 across a single RAID member.
In this way, you'll never loose redundancy.