> rescheduling does not necessarily indicate greater likelihood of default
That is simply untrue. Look, I work for a credit card company, and we have a great deal of experience with evaluating customers and predicting risk of default. A customer choosing to reschedule a loan does not necessarily mean that they WILL go in default, but it is almost certainly an indicator that increases the probability that they will. I am not suggesting you should consider them in arrears (they aren't!) but that you should take payment rescheduling into account as a leading predictor of future default rates on cohorts of loans that have not yet reached default.
> we aim to improve our Statistics page to use a single metric to encompass both ended and current loans...
Depending on how well that metric is designed, that might actually LOSE information.
> ...and allow users to filter the data by date disbursed.
That will definitely be useful. In my experience, graphing key indicators (like default rate, repayment rate, loan profit, number of loans made, and leading indicators of default) by cohort (grouping by start date) is a really good way to view your data to determine whether changes in lending policy are having the intended effects. Since your policies have changed recently, this sort of view may prove very useful.