Having worked with the sort of IDC termination systems commonly used in phone networks, including plenty of 237A disconnect blocks (which is what I imagine this story is about, or something very similar) I would expect them to be steel in order to get the right springiness. Something like copper would not have the right mechanical properties. I would expect the steel to be plated in order to minimise corrosion.
I could speculate that the problem is the contacts that are closed when a test connector is not inserted. The contact pressure is probably a lot less than that between the wires and the IDC terminations, so it's quite conceivable that corrosion could occur there and create a bad joint - which would disappear when testing the line as inserting a test connector would wipe the surfaces clean. I've come across a number of such test points that have been damaged (I speculate by shoving incorrectly sized objects like screwdrivers into them asa bridging tool), and hence the block needed to be jumpered across to bypass these damaged contacts.
For the un-initiated (if they are still reading), the 237A disconnect block has 10 circuits, 2 wires (one pair) each - and it's called "disconnect" because inserting a test probe disconnects the circuit. Connections from (e.g.) the exchange equipment would be connected to one side, and the connections to the subscribers' lines would be connected to the other. Normally, internal contacts pass the circuit from one side to the other, but a test probe can be inserted which a) opens the contacts and thus splits the circuit, and b) gives access to the connections of one side. So at the exchange, you could insert the test probe, disconnecting all the wiring from the exchange, and thus test the exchange port. Insert the test probe the other way round, and you disconnect the line from the exchange and can test the wiring downstream.
There will be multiple points where this is possible. In the exchange itself, the exchange ports are terminated onto banks of them, and there's a 'kin big frame across which jumper wires can be routed to similar banks of connections that feed the multi-pair cables out to the green boxes. Then in the green boxes, the same thing happens on a smaller scale to route from the big multi-pair cables to smaller ones that feed up to the top of poles or to underground joints. And so, if the connector blocks are "defective", there's multiple points where this can cause a problem with services.
Biting the hand that feeds IT
