Abstract On 2019 August 14, the LIGO and Virgo detectors observed GW190814, a gravitational-wave signal originating from the merger of a black hole (BH) with a compact object. GW190814's compact-binary source is atypical both in its highly asymmetric masses and in its lower-mass component lying between the heaviest known neutron star (NS) and lightest known BH in a compact-object binary. If formed through isolated binary evolution, the mass of the secondary is indicative of its mass at birth. We examine the formation of such systems through isolated binary evolution across a suite of assumptions encapsulating many physical uncertainties in massive-star binary evolution. We update how mass loss is implemented for the neutronization process during the collapse of the proto-compact object to eliminate artificial gaps in the mass spectrum at the transition between NSs and BHs. We find it challenging for population modeling to match the empirical rate of GW190814-like systems while simultaneously being consistent with the rates of other compact binary populations inferred from gravitational-wave observations. Nonetheless, the formation of GW190814-like systems at any measurable rate requires a supernova engine model that acts on longer timescales such that the proto-compact object can undergo substantial accretion immediately prior to explosion, hinting that if GW190814 is the result of massive-star binary evolution, the mass gap between NSs and BHs may be narrower or nonexistent.