MESFET vs diode mixers

[radar_macgyver]

I've been staring at the data sheets for MESFET mixers (example: Mini circuits HJK-372H+) for a while. They have excellent compression characteristics (+14 dBm P1dB), decent LO to RF isolation (> 45 dB at my frequency of interest) and the noise figure should equal the conversion loss (7.7 dB) since it's a passive device. All of these, particularly compression, look better than pretty much all of MCL's range of diode mixers, so there must be some reason diode mixers are still being sold. In other words, where's the catch? It can't be the high LO drive, a representative Level 17 diode mixer (SYM-30DHW+) is worse than the MESFET device in every metric, with the exception of LO port VSWR. What am I missing?

[CJay]

HJK-372 is almost twice the price of the SYM-30, the application might not need the extra performance thus wouldn't justify the extra cost.

Given that they are using MESFETs I also wonder if perhaps they're more static sensitive than a diode mixer and prone to damage?