I think I have created a workaround for this one, or this might at least give you an additional idea.
If you create an Isfirstoccurrence based on the date, then you can find the first time this date occurs on your dimension:
Once we have that then we can Rank based on the First occurrence. The 4th argument in the Rank is a Bool, which is what our isfirstoccurrence is going to be. We also need an If statement here to not return a NaN value. My Formula in my rank line item looks like this: IF Isfirstoccurrence THEN RANK(Survey Date, ASCENDING, SEQUENTIAL, Isfirstoccurrence) ELSE 0
Then, I create a Line/module that has no dimensions except for Day. My Formula is going to read: Rank[SUM: Survey Date] . This will bring in the Ranked value to the individual date so that we can reference it by the rest of the items using the same date.
And Finally, I add my Final Rank. The formula is looking at the prior line item and looking up the date of that record. The formula reads: Time to Data[LOOKUP: Survey Date]
As you can see, I now have all of my items ranked sequentially and have no numbers that are being skipped.