As a zero bias detector, up to 1 GHz I prefer BAT62 over 1N5711. Capacitance is 0.65 pF max, while for 1n5711 it 2 pF (I thought). When you make sure DC load is 1 Mohm or more, you retrieve over 80% of detected EMK.
Do not use BAT62 in a mixer or high bias situation as the intrinsic resistance is rather high. This intrinsic resistance (200 Ohms) limits its usable frequency despite the low capacitance. When voltage to be detected is low (max 2V RF amplitude), a better choice is BAT15 (cap < 0.35pF, so it is closer to 1N21). Though BAT15 is a mixer diode (12 GHz), video resistance is not that high (100 kOhm).
When you make sure the load resistance >> video resistance (that is the diode resistance at zero bias), sensitivity is as good as sensitivity of special detector diodes up to many GHz. When you match the diode, sensitivity is really high.
The advantage of the special very low barrier detector diodes is that you can use them at zero bias with moderate load. So you don't need a high impedance input. Of course you can bias a normal diode to get low video resistance, but than DC output measurement becomes more difficult because of thermal aspects.