A new preprint (arXiv:2604.15430, not yet peer-reviewed) reports the first three-dimensional general relativistic magnetohydrodynamic (GRMHD) simulation of sustained accretion onto a horizonless compact object rather than a classical black hole. Using the Joshi-Malafarina-Narayan (JMN-1) spacetime - a solution that emerges from gravitational collapse with anisotropic pressure and chosen compactness so the central singularity is null - the team evolved magnetized plasma until it settled into a magnetically arrested disk (MAD) state. For parameters modeling the low-luminosity active galactic nucleus M87*, the simulated accretion rate reached approximately (3.0 ± 0.5)×10^{-6} Eddington rates, statistically indistinguishable from standard Schwarzschild black hole runs. Synthetic 230 GHz images produced via polarized general relativistic radiative transfer reproduce the ring-like structure and overall brightness seen by the Event Horizon Telescope. Yet a decisive difference appears: measurable emission originates from inside the apparent shadow, sourced from plasma orbiting mere gravitational radii from the naked singularity - a region that would lie permanently behind an event horizon in classical black holes. This preprint distinguishes itself from earlier analytic shadow calculations by capturing full 3D turbulence, magnetic reconnection, and mass inflow that previous two-dimensional or force-free models could not sustain without rapid outflows or accumulation. What much of the existing EHT coverage missed, including the landmark 2019 ApJL papers (arXiv:1906.11243), is that current 230 GHz data sit within the degeneracy region where several horizonless mimickers produce statistically compatible shadows once integrated through the same interstellar scattering and instrumental effects. A related 2020 study on exotic compact objects (arXiv:2002.01938) already warned that photon rings alone cannot rule out alternatives; the present work supplies the missing dynamical accretion history and identifies an explicit, albeit faint, internal emission feature. The finding carries implications that extend beyond one specific mimicker. Decades of theoretical work on the information-loss paradox, firewalls, and singularity resolution have repeatedly produced horizonless objects whose external multipoles can be tuned to match Kerr metrics at large distances. This simulation demonstrates that such objects can also support the exact MAD accretion regime favored by EHT collaboration analyses for explaining M87* jet power. If confirmed, the presence of even a few percent flux inside the shadow would require rethinking whether the dark central patch is truly caused by an event horizon or by extreme redshift and lensing near a timelike singularity. Limitations are important to note. The JMN-1 solution assumes particular anisotropic pressure profiles during collapse whose astrophysical realization remains speculative. Results are reported for one compactness value; different choices can shift the photon sphere and alter photon escape probability. Like all GRMHD work, the output depends on numerical floor prescriptions, grid resolution, and radiative-transfer approximations. No real astrophysical sample exists - this is a single numerical experiment contrasted against one reference Schwarzschild run. Nonetheless, the authors correctly observe that the predicted inner glow lies within the dynamic range expected for next-generation extensions of the EHT array (ngEHT), which should reach the necessary contrast ratios within the next decade. The study therefore supplies a concrete, falsifiable discriminant between the standard black-hole paradigm and a class of alternatives that until now lacked realistic accretion signatures. In synthesizing these threads, the work reveals how tightly current EHT images still cling to the no-hair assumption while simultaneously exposing a narrow observational window that could force a revision of our understanding of strong-field gravity, singularities, and the fate of infalling matter.