A fair enough criticism. This was operationalized by the "Top" column, with the 1% picks being in the historical top 1% of expected yield, and so on. However, in reality, anything in the top 10% I've found performs about equally. Honestly, I'm considering dropping the "Top" column anyway.
Just as you are unbothered by the loan purpose going undisplayed, so should you be unbothered by the absence of interest rate and expected yield. The model already compares the interest rate vs. the risk, and identifies those in the historical top 10% of risk vs. reward. If loan quality was consistently horrible across the platform, the model would not identify any picks (and you could unsubscribe).
That said, I will take under consideration the inclusion of the stated interest rate / expected default (as LC does, but with my own model). Eager to hear from others about this.