OK, after a bit more work with the spreadsheet, I have now shifted the relative placement of my curves up to simulate reducing the exposure (if you remember I had overexposed the stouffer step wedge thereby not getting into the toe region).
This shifting permitted me to create the toe for each curve. I already had two points in the toes on the curves, the FB+F and the density of one step from an X-Rite cal film having a Tx_density of 3.80.

Now having done this, I notice that I probably didn't develop for long enough. Given I was using Xtol 1+2, even though I extended out to nearly 32min@20degC, it looks like I needed to take it out to perhaps >50 minutes in order to get out to N+3 (avg. gradient =1.0) on the curves. Does that sound reasonable ? If so, then to save me exposing yet another film, can I assume that it is now OK for me to extrapolate my existing curve of Zone System [N] vs. development time[min] (bottom left of the 4 curves) out to say t=50minutes ?

One thing which makes me weary of doing this is that Ralph's example had him reaching N+3 and his corresponding curve is much more linear above N=0 rather than concave down (like mine). Now of course he used Tmax100 in ID-11 (I used HP5+ in tol 1+2) which could account for the different shapes, but I'm not familiar enough with the shapes to know if my shape is OK and whether I can extrapolate it or not.

My test summary:
Summary.png

Ralph's Test Summary:
Ralphs Summary.png

My Family of Curves:
Family of Curves.png

My spreadsheet:
FilmTestEvaluation #002b of Ilford HP5plus 120 in XTOL 1+2.xls