The radial spatial distribution of low-mass satellites around a Milky Way (MW)-like host is an important benchmark for simulations of small-scale structure. The distribution is sensitive to the disruption of subhalos by the central disk and can indicate whether the disruption observed in simulations of MW analogs is artificial (i.e., numeric) or physical in origin. We consider a sample of 12 well-surveyed satellite systems of MW-like hosts in the Local Volume that are complete to $M_V<-9$ and within 150 projected kpc. We investigate the radial distribution of satellites and compare with $\Lambda$CDM cosmological simulations, including big-box cosmological simulations and high resolution zoom in simulations of MW sized halos. We find that the observed satellites are significantly more centrally concentrated than the simulated systems. Several of the observed hosts, including the MW, are $\sim2\sigma$ outliers relative to the simulated hosts in being too concentrated, while none of the observed hosts are less centrally concentrated than the simulations. This result is robust to different ways of measuring the radial concentration. We find that this discrepancy is more significant for bright, $M_V<-12$ satellites, suggestive that this is not the result of observational incompleteness. We argue that the discrepancy is possibly due to artificial disruption in the simulations, but, if so, this has important ramifications for what stellar to halo mass relation is allowed in the low-mass regime by the observed abundance of satellites.