Data on UDs?
The biggest concern is the drop ratios – what is the common/masterwork/rare ratio of dyes that you get? I’m perfectly happy to assume that all dyes of the same tier are equally common, but nailing down the ratios of each tier is a big deal and would have a pretty big effect on your expected value from cracking a dye.
I’ve read 60/30/10 before, which passes the smell test – but I’d want a sample size of ~1000 dyes to be confident in my result, and the last time I ran the numbers I concluded that cracking unidentified dyes was a money losing proposition (at those numbers), so I didn’t pursue it further.
I’ll take another look when spidy comes back up, if the outlook is at least neutral I’ll think about tossing a buy order down for a thousand dyes.