Hi,
In my opinion, for instance, if you create a relationship between the DATEKey column from the DATE table (20220101 -> number data type or text data type) and the OrderDateKey column from the Sales table (20220101 -> number data type or text data type), this will not work if you do not mark as a date table.
However, in this case, because you created a relationship between the Date column (Date data type) from the Date table and the OrderDate column (Date data type) from the Sales table, in most cases, it will work even without doing "mark as a date table".
This is especially when using time intelligent DAX function. In most cases, when you see time intelligence function, most of them refers to the Date column (Date data type) in the Date table. This creates something like "removing all filters in the Date table and then recreate filters as you wish". However, if the relationship is using datekey column (text data type or number data type), then time intelligence DAX function will not create "removing all filters in the Date table and then recreate filters as you wish" if you do not mark as a date table.
Please try adding one more column in the Date table and Sales table.
Date[DateKeyColumn] and Sales[OrderDateKeyColumn] those look something like, 20220101 (number data type or text data type), and then create a relationship based on the two newly created columns. And unmark as a date table. -> I think this will not work when you use time intelligence DAX function.
Of course rest of measures will work those are not using time intelligence DAX function.
I hope this helps.
