Here's how I solved my problem.
I needed to set certain text and error conditions depending on the underlying data. The logic was fairly complex. Qlik's load script did not have the flexibility to easily do this analysis. I therefore took it out of Qlik Sense and wrote some Python code. The Python code created a new file of error conditions, including the name of the data source (the key), the error condition text, and the error condition itself.
I then read in this error condition file as a new table, assigning colors to the error condition. I used the table chart in Qlik Sense to display my results. I ordered the table by error condition (most severe first) and colored it using the error condition colors. At the bottom of the sheet I added a chart that updates when the user selects a data stream.
I could have had different tables for the error conditions, or even some kind of selection box, but it turns out the table works well.
As it turned out, the problem here was a data processing problem, once I moved to Python, it was all OK.