FG readings with my hydrometer (I double checked calibration) 1.025 and 1.024 after adjusted for temperature. None of them come that close, off by about 5 points across the board. This is similar to what I saw before.
Thanks for posting this, Tom. The problem you're seeing is due to the fact that the OG range used in the initial data set was 8.8-24.3°P, and outside that range the correlation falls off pretty quickly. I'm seeing this with all the very high- and low-gravity data people have sent in. The other major issue I want to address is that I didn't test any worts with (real) attenuations less than 59%, which I thought was a reasonable floor for typical beers - apparently I was wrong about that, since I've gotten several submissions with RDFs in the 40s. There's going to be a 3.0 release in the next week or two that will hopefully address both issues.
FWIW, the new correlation I'm playing around with (as of now) puts the FGs at:
A: 1.022
B: 1.021
Which still isn't "good enough" IMO, but it's getting there. How confident are you in the 25.5°Bx value? If you'd like I could incorporate it into the new dataset.
Also, how did you cool the sample? I've noticed that if I don't chill it in a sealed container evaporation can increase the gravity significantly. It might not explain the discrepancy in the spreadsheet, but I thought it was worth mentioning.