Rolling Cumulative Historic Date Filter

The way I normally solve such problems is to use a time intelligence table that contains all the complete sets of dates for 30, 60, 90 etc. Doing this will create date duplicates, so you need to set it as many to 1 to the date table and turn on bidirectional cross filters. 
I always create such a table using union queries inside "new table" using DAX. You could do it other ways. 

The following is illustrative only

DateUnionTable =
VAR Last30Days =
    SELECTCOLUMNS(
        FILTER(
            'Calendar',
            'Calendar'[Date] >= TODAY() - 29 && 'Calendar'[Date] <= TODAY()
        ),
        "Date", 'Calendar'[Date],
        "Period", "Last 30 Days"
    )
VAR Last60Days =
    SELECTCOLUMNS(
        FILTER(
            'Calendar',
            'Calendar'[Date] >= TODAY() - 59 && 'Calendar'[Date] <= TODAY()
        ),
        "Date", 'Calendar'[Date],
        "Period", "Last 60 Days"
    )
VAR Last90Days =
    SELECTCOLUMNS(
        FILTER(
            'Calendar',
            'Calendar'[Date] >= TODAY() - 89 && 'Calendar'[Date] <= TODAY()
        ),
        "Date", 'Calendar'[Date],
        "Period", "Last 90 Days"
    )
VAR Combined =
    UNION(
        Last30Days,
        Last60Days,
        Last90Days
    )
RETURN
    Combined