I don't think that was a temp comparison or that it had anything to do with temp accuracy. It's his attempt to convert the I2C to sensibus. He's trying to understand how the 75 is read and adapt the 35 to that. Basically he has it down to 0.0-0.1F and 0.0-0.1% RH. IOW, what the 75 would report as say 85.4F, the 35 would read it as 85.3F or 85.4F. So there is probably some 0.05F error in there which he is trying to fine tune. Personally, a 0.0-0.1F conversion error on top of the 0.1-0.2F sensor accuracy is good enough for me.