Remember that a jet engine produces thrust by accelerating a relatively small (compared to a propeller) mass of air by a relatively large amount. EPR is trying to quantify this by representing on instruments the ratio of tailpipe pressure to inlet pressure.
I think when Silberfuchs says EPR is not a linear indication of thrust, he means that 2.0 epr is not twice as much as 1.0 epr. It is actually alot more. I would say it is almost a logarithmic scale.