The LIGO/Virgo Scientific Collaboration (LVC) recently announced the detection of a compact object binary merger, GW190425, with a total mass of 3.4(-0.1)(+0..3) M-circle dot and individual component masses in the range of about 1.1-2.5 M-circle dot. If the constituent compact objects are neutron stars, then the total mass is five standard deviations higher than the mean of 2.66 +/- 0.12 M-circle dot for Galactic binary neutron stars. LVC suggests that the nondetection of such massive binary neutron star (BNS) systems in the Galaxy is due to a selection effect. However, we are unable to reconcile the inferred formation efficiency from the reported merger rate, R-GW190425 = 40(-390)(+1050) yr(-1) Gpc(-3), with predictions from our own study for fast-merging BNS systems. Moreover, the comparable merger rates of GW190425 and GW170817 are possibly in tension with our results for two reasons: (i) more massive systems are expected to have lower formation rates, and (ii) fast-merging channels should constitute less than or similar to 10% of the total BNS systems if case BB unstable mass transfer is permitted to take place as a formation pathway. We argue that, to account for the high merger rate of GW190425 as a BNS system, (i) our understanding of NS formation in supernova explosions must be revisited, or (ii) more massive NSs must be preferentially born with either very weak or very high magnetic fields so that they would be undetectable in the radio surveys. Perhaps the detected massive NSs in NS-white dwarf binaries are our clues to the formation path of GW190425 systems.