You can't use the cut angles of the appendix to place the rails BR3/BR4 and BR1/BR2. You need to use the check dimensions p34 of the book. That is, to my mind, why your BR10 does not fit on your CAD drawings.
Why that?
Considering the fact that the author of the book considered that we will draw the layout by hand on a table, using the check dimensions given on page 34 of the book, the exact position angles of the rails are not given in the book. But considering the fact that cutting a tube can not, for most of us, be accurate at a 10th of a degree, the cut angles in the appendix should have been "rounded" to the closest entire value, which, in this respect, makes sense.
Therefore, the cut angles can not be used to place the rails on a CAD layout or chassis modelling, because they hold tiny mistakes, consequences of the rounding of the values in the appendix. These tiny mistakes are acceptable on the cut of the section of a 25mm square box section tube. But if you place the rails according to these angles, theses tiny mistakes are actually multiplied by the lengths of the rails, and at the other end of the tube, the check dimensions are not respected. I hope it makes sense.
The book's datas are not incorrect, they are just not supposed to be used directly with such an accurate tools like a CAD software.
You can check the layout I've drawn (and used on my build table). I wrote the explanations about this necessary correction in English at the right of the layout, and wrote on the tubes, the exact cut angles you can use to properly place your rails in your CAD drawing.
You can open the documents here:
https://drive.google.com/folderview?...Iz&usp=sharing
Or, to make it shorter, consider the cut angles of BR1/BR2 as being 15.667° and not 16°, and angles for BR3/BR4 (and BR10) as being 9.6° and not 10°. With these angles, the check dimensions are respected, and your BR10 should fit. In the same way, BR8/BR9 should have a cut angle of 74.333° at the front end instead of 74°.
These are personal calculations, to me, they are correct, but you might have more accurate tools than me to get even more accurate values considering what's explained above. I don't pretend I've redrawn and redesign the layout and that it is what should be considered as the divine truth
Making a 3D CAD model is a good way, to my mind, to get familiar with the chassis design, if you have the time and skills (and it looks like you do
), but if you just need to have a CAD model of the chassis, you can spare yourself the time of drawing it, as there is one professionally done by madis on Solidworks which is really good and freely shared here:
http://grabcad.com/library/haynes-roadster-chassis
Hope this will help