We spent time here in the shop doing basic strength tests, taking a length of maple that was 1"x 1" and 12" long, then hanging a 10 lb weight from the end and measuring the deflection. There was a measurable difference between flat and quarter sawn of approximately 1/10th of an inch. This was done at least 4 times with different "samples" of rock maple, and each showed the same (within 1000th of an inch) results. May not be fully scientific, but it seems a reasonable result.
Well, to be fair, the study I was quoting is a little dated. But I have to say that the tests performed were extensive. The resulting publication runs to over 350 pages with 30+ pages of tables showing consolidated test results. The tables cover things like volumetric, tangential and radial shrinkage, work in bending in.lb/ per cu.in., impact bending, modulus of elasticity, compression stresses, hardness, cleavage (splitting strength, nothing to do with v-cut dresses), tension perpendicular to grain, and on and on. It's been my wood bible for quite a few decades. I have never found another publication like it.
For a particular species of maple the tests showed that there was a bending differential between the radial and tangential planes a tad less than 10%, but the wood was about 10% harder on the tangential surface than the radial one (other species were similar, but not identical). Shear strength parallel to the grain was about 5% better on the tangential boards. However, the results were not identical when comparing air-dried to kiln dried wood. So one has to be careful about the use of the term "strength". And a bass neck is subject to compression forces as well as bending forces and so a generalization about tangential vs radial strength gets rather dodgy. There are other factors that are more important for neck construction - density, drying technique and thoroughness, case hardening, tooling, grain orientation (runout), type of glue used, etc. etc. etc.
I am not doubting your test results. I'm just saying there is more to the term "strength" than the amount a bend a static weight induces in a piece of wood.
Just for fun, try a little experiment using the same setup you describe above. But instead of comparing the radial flex to the tangential flex, turn the test piece over, so what was the bottom face is now on top. It is likely you will find that the wood does not flex to the same degree. The difference is small, but since you are measuring to 1000th of an inch, you should be able to see it. Damn trees - they are so inconsistent.