It's understandable. That's why I was thinking if the data was already refined to only DT1! inclusions/exclusions, it'd be a lot less data to parse (presumably). But then there's still the issue of thousands of users on DT2 -1, that in all honesty I don't really care about either. There's probably a logical way to parse the data without it taking too long though I imagine, I just wouldn't know how.
I think it's best if I scrape the total DT-strength for all users on DT1 and DT2, instead of just DT2. If it's just DT2, the gaps can either mean the user is on DT1 that month, or the user has a negative DT2-strength. I don't think that gives an accurate overview, so it's better to show the total "DT-strength". What do you think?