@majorp93 no, these updates should not change anything on the MTU behavior. From your screenshots and ip a output, I assume eth4 is the only link used, than your vlan1060 network is indeed your management network. I guess with some confidence that:
- eth4 is your actual NIC
- xenbr4 is the pool-wide network without vlan using eth4
- xapi1 is the bridge for vlan1060
If that's right indeed on your host side, the MTUs are properly set.
As Gaël said, jumbo on management network is not officially supported, because we always end up in situation similar to yours, where something stops working for "some" reason 
But as you said, when everything is setup properly, it does work.
I have test hosts at home up to date but no similar setup to yours for SR. I tried with a pool-wide and a pool-wide + vlan as management with 9000 and the ping -M do -s 8972 does work fine in both cases, so nothing I can see here.
When it fails, what do you see? Message too long? mtu=1500 ? something else?
Do you have multiple hosts in that pool? Can you try that between the hosts as well and not toward the storage systems? I would also check on ovs with ovs-vsctl list interface that each of eth4, xennbr4 and xapi1 report 9000 mtu.
I know you said it was properly setup everywhere, but I would still be tempted to think there is a setup issue somewhere.
