> Each data has to be written on the host itself and on the two remote ones so you are going to have 1000 mbps / 2 (external replicas ) / 8 (bit/bytes) = a max of 62.5 MB/s sustained throughput shared between all the VMs and this ignoring all the overheads.
> In practice it will be much less ending in a barely usable environment.

I am currently in this type of setup (replica 2 arbiter 1 on 1 gbit/s  both storage and oVirt connectivity)  and the maximum write (sequential) speed I can reach is aprox 89 MB/s.
I guess with replica 3 the maximum will be as Simone stated.
My maximum read speed is aprox 500 MB/s  despite that local NVMe is capable of 1.3 GB/s and I guess  FUSE  is not using  the  local brick no matter  I have set it up with cluster.use-local  (or whatever  it was called).

If possible , get a 10 gbit/s connectivity or multiple  gbit interfaces  ( as  I'm planing to do in the nearest future).

Best Regards,
Strahil Nikolov