The short would be that real time replication of metadata is really hard and leads to a lot of overhead. I know they are planning on building something that allows you to migrate your data to a new satellite, but that wouldn’t solve this problem.
I don’t think you really need it to use the data from a different location though. Latency between uplink and satellite doesn’t matter all that much as transfers happen directly to the node anyway.
As for letting nodes do audits and repair. I wrote a suggestion to distribute audits a while ago. I wanted to write one for repair too, but never really got around to it.