We present the results of the second year of exoplanet candidate host speckle observations from the SOAR TESS survey. We find 89 of the 589 newly observed TESS planet candidate hosts have companions within 3″, resulting in light-curve dilution, that, if not accounted for, leads to underestimated planetary radii. We combined these observations with those from Paper I to search for evidence of the impact binary stars have on planetary systems. Removing the one-quarter of the targets observed identified as false-positive planet detections, we find that transiting planets are suppressed by nearly a factor of seven in close solar-type binaries, nearly twice the suppression previously reported. The result on planet occurrence rates that are based on magnitude-limited surveys is an overestimation by a factor of two if binary suppression is not taken into account. We also find tentative evidence for similar close binary suppression of planets in M-dwarf systems. Last, we find that the high rates of widely separated companions to hot Jupiter hosts previously reported was likely a result of false-positive contamination in our sample.