The LM317 needs ~2-3V drop between IN and OUT to regulate properly. The OUT pin stays 1.25V higher than the ADJ pin when in regulation.
You can try connecting the ADJ pin to ground through a 10 ohm resistor, see if the output drops to 1.25V. The voltage across that 10 ohm resistor should not be more than 0.1VDC (~ 10mA current = ADJ pin current + voltage divider current.) If these things are true, the LM317 is at least basically functional, thogh it might be a little leaky.
If the part reads a few ohms between IN and OUT (+ to IN) with power removed and the capacitors discharged (to ensure a correct resistance reading,) then find out why. Resistance readings in-circuit can be misleading at times, but a few ohms is pretty definitely bad in this case (barring an external component across the LM317 that's shorted)
Normally the ADJ pin current does not exceed 50 microamps, if significantly more than this with 2-3V across in-out, then the regulattor is bad.
They are probably switching resistor values to ground (with NPN transistors) in order to adjust the output voltage, this is very easy to do with an LM317. Driving the ADJ pin to ground will yield an output of 1.25V (essentially off for an LNB.) 2 transistors would be enough to do so using 2 series resistors for R2, one transistor shorts out the bottom resistor, the other shorts out both resistors. Both off=18V, 1 on (bottom R shorted)=13V, other on (both R's shorted)=1.25V (off.) Vout=1.25 * (1 + R2/R1 ) R2 goes from ADJ-ground, R1 from OUT-ADJ. R1 is ~120-240 ohms to ensure leakage current and ADJ pin current do not affect output regulation.
A diode is often connected across the LM317 in-out pins, K to in pin, A to out pin, to shunt reverse current around the regulator if the input suddenly drops or a spike comes down the cable.