Johnson Su Transformation Parameter Discrepancy

[#1]

Hello from a novice user. Our organization recently updated to JMP 17 from JMP 15. The typical extent of my analysis is evaluating data sets for normality and calculating tolerance intervals. It came to my attention that using the "Enable Legacy Fitters" tool and utilizing a Johnson Su transformation will yield different transformation parameters than using the new "Fit Johnson," even if Johnson Su results is the best fit. Additionally, forcing a Johnson Su transformation can situationally result in a normally distributed data set whereas using the new transformation on the same distribution will result in a non-normal data set.

I have attached an example of this and would appreciate any insight as to why this may happen and how I can better utilize the Johnson Transformation function.

[#2]

I would also like any guidance on this inquiry.  I noticed the same thing and wondering if an explanation can be provided.