Dynamic Date Filter

Following is my suggestion.

I created a small sample consisting of 2 tables. Please don't be confused by the German date format, it's read dd.mm.yyyy.

First table contains the campaings:

Second table contains the data to be analyzed, in my sample a simple sales table:

You can create a simple slicer with the campaings, as you already did.

In your measures you can add a test condition that checks whether there's excatly ONE campaign selected in your slicer. The trick is to use the HASONEVALUE() function. If it returns TRUE (meaning there is one campaign selected), narrow the results of your calculation with the start and end date of the selected campaign. Otherwhise just ignore the dates. A measure that sums up the Amounts looks like this:

SalesAmount = IF(
    HASONEVALUE(Campaigns[CapaignId]),
    CALCULATE(
        SUM(Sales[Amount]),
        Sales[SalesDate] >= VALUES(Campaigns[StartDate]),
        Sales[SalesDate] <= VALUES(Campaigns[EndDate])),
    SUM(Sales[Amount]))

The results with different slicer selections:
1. No slicer selection:

2. "Campaign 1" selected:

3. "Campaign 2" selected:

This approach works if the user can only select one campaign in the slicer. If multiple selection is allowed this sample has to be enhanced.

Hope this helps!

JJ