In the RIAA specification, frequencies from 20 to around 500Hz are boosted with a slope of about 6db/octave. At 500Hz to 2kHz, there is a plateau; those frequencies are not boosted. Then from 2kHz on up, the boost resumes at the same slope. If you are accurate in placing the peak, then maybe there is something askew with the RIAA (there may be no plateau for frequencies between 500Hz and 2kHz due to a faulty IC or whatever), but you have also to take into account that your other LPs sound good. That really suggests that either you are particularly sensitive to the instruments and the voice on those few LPs that bother you, or that they are otherwise the source of the problem. Others might also suggest you check all aspects of cartridge alignment. By the way, not many male vocalists (except castratos, maybe) can get up to 2kHz or even 1khz. So maybe your issue is lower down in frequency than you estimate.
To amplify on what I wrote above, Mijo is correct roughly speaking to say that as internal R goes up, gain using a current driven stage goes down, but this is because the voltage output divided by the internal R gives a rough approximation of current output of the cartridge. So as R goes up, if you hold voltage output constant, current output goes down. The other factor is the actual input impedance of the current driven stage. An ideal current driven stage would have zero input impedance. But that is not feasible; if input Z is zero then the signal will go to ground. That is the equivalent of a mute switch. So, all of these stages have some finite input impedance, and manufacturers are not good at revealing numbers. The actual input impedance of the stage will have a big effect on the efficiency with which the stage is actually driven by current. Typically, the impedance is between 1-2 ohms to as high as 10-20 ohms. If you mate an LOMC with an internal R of 2 ohms to a "current drive" stage with an input Z of 20 ohms, then you get voltage drive, depending upon the sensitivity of the input to voltage. This is why I hate the term "transimpedance", which implies that impedance doesn’t matter. It most certainly does make a big difference in the success or failure of these cartridge/phono stage matches.