They're all expensive and they all want to lock you in, because they're businesses. GPU's are in extremely high demand now so that's just how it is. And they're complex because different customers have different needs, so that's inherent.
I don't see how any of your complaints are anything a for-profit enterprise cloud provider could change and still survive as a business.
There is only one vendor (for anything scientific/ML anyway) and you have to buy specific (overpriced) models from them to be compliant with the license. So zero negotiating power, and everything is proprietary so you cant build any software/support edge.
You cant oversell a GPU like you can CPU/memory, and most of the computation happens on the chip so no one cares how good the rest of your stack is.
And finally you are in a neverending battle with crypto miners, which would never be profitable on cloud so they have no intention of paying you.
So the (slight) differentiations are.
AWS - gpudirect and high bandwidth (but also high latency)
GCP - 16 gpus on one machine
Azure/OCI - mellanox
Try Lambda Labs, there's usually more availability, and more importantly you get some visibility into currently available capacity.
You are welcome to install MySQL or elasticsearch on another VPS outside of AWS and use it. But that's not what customers want.
The comparison here is apples & oranges
I'm all for bagging in Oracle in general, yet it's undeniable this is one area where they're leagues ahead and winning.