That most orthopedic tests haven’t been proven definitive in diagnosing some or most musculoskeletal conditions is a given. In fact even those tests with good inter-reliability suffer with short comings in validity. Laslett, Bogduk et al did a systematic review in 2010 to determine whether tests could accurately determine a disc vs. facet vs. an SI joint and found the answer to be “generally NO”. However most experts suggest a barrage of tests (with the majority negative (or positive)) give an acceptable odds-ratio as to the likelihood of involvement. This is more so the case with the SI joint vs. facet or disc.
The above picture from SI-Bone suggests a clinical testing procedure with the (5) most diagnostically-relevant SI tests. IF (4) of the (5) are positive there is a 75% likelihood the SI may be involved. The opposite is true as well ((4) of (5) negative). However other studies suggest the MOST relevant tests are the THIGH TRUST and distraction/compression. These (2) tests tended to give the strongest indications.