the tar/coke theory is utter crap. its not possible. it would take a huge lump of crap to block a valve open, and even if it stuck a valve open you would only lose compression on 1 cylinder.
Crank sensors are tested according to mechanics, all gives right readings. With old box it gives the crank sensor/Ecu fault code. With the new one no fault codes. Fuel pressure is ok according to mechanic, also fuel back pressure on return pipes.
to fully test the crank sensor you need an oscilloscope and have the flywheel spinning with ignition switched on (easiest way is to use starter motor with sparkplugs removed).
a faulty crank sensor usually wont trigger a fault code because the ecu can only determine if the signal is missing or obviously wrong (out of range), most crank sensor faults dont cause an error to be recorded by the ecu.
in my opinon the crank sensor must be changed. they should use some common sense rather than a diagnostic.
The compression test is the key, but according to workshop they haven't long enough tools to get it done (Bosch and Alfa workshop). -> they want to dismantle top to check the problem. In my opinion it sounds a bit strange....
they are obviously crap, all you need to do is stick some insulating tape around the tester and it works fine, i've done this many times, do not
let them take the head off unless a compression test has shown that there is a loss of compression.
i really think it is the crank sensor, its another classic case that ticks every box, at least try that first because it is cheap and easy to do.