Ciao Giampiero, I can add my 2 cents from my personal experience.
The great advantage that Delphix provides it's in the capability to strengthen and speed up the upgrade process.
In real life scenarios, many time a DB upgrade is not executed because of the many different impacts this task can have on a prod environment, where on top of the DB there are one or many applications using the DB to consume and/or persist data.
In fact, an application is made of processes, functions, events that can be heavily affected by the upgrade version of the DB.
For example, in best case, since you're upgrading you DB, you would expect to improve the application efficiency by exploiting new capabilities provided by the upgrade.
In worse case, you could experience difficulties and stops during the upgrade process, post upgrade performance issues, lack of functionalities, etc.
All these cases suggest to plan properly an upgrade task in order to minimise risk and production outages.
This means to test and reiterate an upgrade process until it's fine tuned and proven. Imagine it as a Software Development Project and its life cycle. It is actually.
Hence, the more you can test it, the better and faster you achieve your goal.
Delphix provides exactly these benefits: it allows more tests in less elapsed time.
In some enterprise application projects I've worked on, a DB upgrade project required weeks or months. With Delphix the same result can be achieved in days.
Back to your comments, yes the upgrade process isn't changing really. You can save a lot of work of a DBA and the Application team, because you de-risk the upgrade, reducing to zero the unprevented, unexpected issues and the related firefightings that happen in prod.
Hope this helps
Ugo