Thanks for your insightful reply Michael!

1) I am fairly sure that the problem is not ESD related since ESD precautions were taken. The packaging was a little suspect since it was not the regular ESD safe packaging I am used to. That is one of the reasons I built the 89C2051 circuit, to test the FET when it comes out of the package.

Coming to think about it, ESD damage may be possible if the earth of the soldering iron is not as it should be. Will check.

2)In this case the white prototype board is innocent! :-) The circuit in question was built on PCB and your impression is correct, the 2051 circuit was built on prototype board. No soldering required... Soldering iron earth fault could maybe explain that too.

3) I am certain that the circuit on the PCB is correct since I replaced the FET directly with the BC517 darlington transistor (increased the base resistor to 10k though) and the circuit works fine now.