Eh?
All the push wire connections I have seen do not exert much in the way of a force on the normal to the surface of contact; it's just a spring pressing against the wire, with the shape of the spring end forming a barb pointing in such a way as to prevent withdrawal of the wire.
Compare this to a screw terminal, which will have a high jointing force, and a crimp which will also have a high jointing force.
The result of this will be that a push wire connector will have only a fraction of the contact area, and only a fraction of the resilience against vibration and shock. Hell, with a large shock, the connector spring could bounce off the wire entirely.
With a small contact area, the remaining conductor surface will be exposed to the atmosphere and oxidise. If there is vibration, shock, or movement caused by thermal expansion and contraction, the low jointing force could result in the conductors moving relative to one another, and the low contact force could then mean that the oxide layer gets in the way of the joint, resulting in increased resistance in the joint.
So I would say that a push wire connector runs a much worse risk of increasing resistance over time compared to a screw terminal, and in turn a screw terminal is not as good as a crimp.
Regardless of what the Regs say, I would be pretty cheesed off if I were to discover that the earth connection for a given circuit was connected through a push wire connector, and boiling mad if I found out it was inaccessible.
Frankly, I like my earth connections to be low resistance, and reliably so over time. When there is a live/earth fault this then means the potentials hopefully divide in a way which keeps your chassis potential low enough to keep you (or your children) alive until the relevant current-balance or overcurrent device trips out.
I just installed in some continental luminaires. The live and neutral terminals were push wire. However the earth connection was a screw terminal. I wonder why that would be?
