On the server, you can associate the reload of a qvw with the end of the successful reload of another.
Let us say that you have an extraction layer with qvws A and B creating QVDs, and transformation layer with QVWs C and D generating QVDs and the final app E.
On the server, on the documents page, reload tab, you set a chain of reloads by checking reload schedule: on event of another (and then choose your document). You can make B run after the end of A, C after B, D after C and E after D.
Hope it helps,
thx for your extended answer!!
Your assumption of the architecture is correct except for the last step we want to load the second layer to the third as a binary because of the speed this delivers. Therefore we need to have the QVW of the second layer itself (with all the new data in it) to be saved... which than can be reloaded into the third (end user) layer!
Thx for yor comments...