If I understand your question, you cannot perform the join because the format of the CSV file (specified in the csvt) is not taken into account.
You should import the CSV with the 'Vector layer' input (as you did). The fields of this file have String type.
Now, let's say you want to join the CSV table with a vector layer on a integer field called "point_ID".
You should thus use a 'Refactor fields' on the field "point_ID" of the CSV table to convert the value from String to Integer and then perform the join. In the Refactor fields tool you have to do 3 things:
- Create a field that has a different name than the field of interest ("point_ID"). Let's call it "point_ID_int". This step is very important.
- The value of the field is set to: to_int("point_ID") (If you want to convert it in the integer format).
- The destination format has to be integer as well.
You also have to refactor the other fields that are important for your calculations.
See picture below for an illustration: