Because I'm lazy and can ask experts instead of trying myself...
I recently obtained a 48G+ because it seemed to be the ultimate product of old HP, subject of Wicke's books, RPN/RPL, build quality, etc.
I was astonished at how long it took to plot the "Woodyard/Kahan integral" over the default interval. Seemed like more than a minute. Then an equally long wait after I changed the interval. Granted, I spent only a few minutes with it, but found the calculator unusable for interactive function analysis based on the output delay. I did a full clear first. Were there some flags I didn't set but should have that would have sped it up?
I have several excellent graphing/CAS programs on a big-screen Mac, but thought it would be cool to mess around on the small screen too using vintage tools.
The 48G+ is no longer with me, but a 50g is not out of reach, and I'm lucky to own an Xpander which I never use because the battery insertion is a major pain and screen contrast atrocious.
Are there any plotting/graphing/function analysis benchmarks comparing 48G to 50G to Xpander? Imagine the performance has improved markedly with new processors.
If you are into plotting then either the new calculator or better yet a software package running on your PC. The 48's are great calculator but the screen resolution is too low for good plotting.
Possible improvement in the 50g. So far I've been impressed in the speed over my old 48SX. However.....
I ran into an odd problem with having the 50g find the extremum of a function today. We were solving the classic problem of optimizing the viewing angle in a theatre where the seats are on a slant. To make the equation more manageable I used two variables for subequations, and then graphed the main equation...
as = (x Sin(25) - 6)^2 + (x Cos(25) + 7)^2
bs = (28 - x Sin(25))^2 + (x Cos(25) + 7)^2
eq = ArcCos((as + bs - 484) / sqrt(4 as bs))
Note: the angle is 25 degrees.
Graphing eq produced the correct graph, but the extr command gave values wherever the initial placement of the cursor was; not anywhere near the maximum.
However, evaluating eq and regraphing the "expanded" function, the extr command gave the correct optimum x = 9.295.
I hate to say it, but my students TI84's did this just fine using Y1, Y2 and Y3, and found the maximum in about 1/5th the time. :(
CHUCK