The "current source" which is connected between +50V and the collector of Q2 is not for protection altough it limits the output current sourcing capablity, too. Its main purpose is to generate high voltage gain. I am sure that term VA means just "voltage amplification".
Voltage at point VA jumps up if Q2 sinks less than the "current source" allows. If Q2 tries to sink more than the current source allows, the voltage at VA drops. The transition slope is very steep. Pair Q1Q2+R2 is voltage controlled emitter follower type current source. Q3 pulls the control voltage at node "from input stage" onto its knees if the voltage over R2 tries to exceed 0,7V.
The "current source" has no internal active pushing capablity, it's a current limiter which probably has exactly the same principle as how Q3 prevents the current of Q2 to be too high.
If you plan to add another "current source" in series with Q2 and to remove R2,R3 and Q3, you would gain nothing. There would still be the same parts and mechanism + an extra transistor in series with Q2. R2, R3 and Q3 would limit the current of that extra transistor.
What is the idea to have thislike protection in an internal stage of an amplifier. Nobody has a possiblity to short its output because there's the actual output stage (=high current capable voltage follower) between this circuit and the speaker connectors? Answer: Some clipping can occur. => The VA stage can get saturated. The saturation decays slowly due the diffusion capacitance in the transistors. Saturated amplifier causes severe distortion longer than a nonsaturated. This is the reason why cheap transistor amplifiers do not allways sound crisp and clear with transient rich music.
here the deepness of the saturation is tried to keep limited.