I have a bunch of quarter bridge 50kg load cells, which I can wire up in single, double, or quadruple connections for different motor thrusts.
The wooden frame just holds things square and true. The load cells push against 1" oak, but metal would work well too.
I have another iteration that uses the INA125 for amplification, but I ended up with more noise than I liked.
I usually calibrate at 2 values (3 if you include zero) but overall the load cells I use are very nicely linear.
I calibrate before each day of testing, but usually only test 1 or 2 motors on it at a time. Calibration changes a little bit over time, but hardly anything really.
I've only had one test with the pressure transducer, but I've been working on designing a BATES grain testing apparatus with the pressure transducer playing a higher part of the testing.
I'm looking at cleaning up the source code and finalizing a PCB design, but it might be after I go down the path of my controlled descent project.