I see it from the outside perspective and I think the future looks pretty simple.
They're going to make the "main thing the main thing"... and not in the Raghu way but triple down in making the vSphere/VSAN/NSX infrastructure abstraction indisputably the underpinning for those large and regulated industry customers. Massive investments with R&D, Support, etc. The customers will get a high quality product and very viable public cloud alternative.
VMware isn't going to cater to those little podunk shops with 100 ESXi hosts - they'll certainly take your money if you want. But the cost of maintaining the relationship isn't going to be worth it to those who think they can continue to get away with paying their grandfathered renewal costs for one of the most important components in their datacenter. If you want a critical product, you're going to pay for it.
The VMware Cloud offerings will be a great bridge for those who don't have the time to refactor all their stuff and are too smart to move to Hyper-V or AHV.
As far as the bullcr-p BU's around the edges and diluting what VMware does, I hope broadcom smokes it
SASE - a complete joke of a play, Velo certainly was a quality product and had its day in the sun. But cobbling WS1 and Carbon black was hilarious then and worse now. Zscaler and Palo own that space. Velo would have no problem finding a suitor
ALB - Roll AVI right into NSX, no reason to have visible separation from code or licensing.
Endpoint/XDR - maybe fuse in XDR with NGAV? I'd still sell Carbon Black off and salvage what you can. Crowdstrike and Cortex (Palo) have ran away with that space.
VDI & Mobile MGMT - Decent stuff, just doesn't fit. Find a buyer, wouldn't ki-l MSFT to have PCOIP and removed Entune competition
Tanzu - probably the steamiest pile of dung. Keep the great work done with Project Pacific and have k8s & container support as a great capability of this wonderful hypervisor subscription and stop trying to monetize a workload form factor on its own island and stick it to Redhat.
The rest of the stuff is garbage, I wouldn't even shop it around and just donate whatever you can to CNCF. Stop trying to monetize kubernetes, just worry about landing the container on your hypervisor.