We present an analysis of archival multi-frequency Very Large Array monitoring data of the two-image gravitational lens system CLASS B1600+434, including the polarization properties at 8.5 GHz. From simulating radio light curves incorporating realistic external variability in image A, we find time delays consistent at 1 $sigma$ for all frequencies and in total flux density and polarization. The delay with the smallest uncertainty (total flux density at 8.5 GHz) is $42.3^{+2.0}_{-1.8}$ (random) $pm 0.5$ (systematic) d (equivalent to $42.3 pm 2.1$ d) whereas combining all delay estimates gives a slightly higher value of $43.6pm1.2$ d. Both values are lower than the previously published radio result and inconsistent with that found in the optical. $H_0$ determination is difficult due to the complicated lensing mass and the lack of constraints provided by only two images. However, analysis of archival Very Long Baseline Interferometry data reveals jets in this system for the first time, the orientations of which provide model constraints. In addition, extremely sensitive maps made from combining all the monitoring data reveal faint emission on one side of the lensing galaxy which we speculate might be the result of a naked-cusp lensing configuration. Finally, we find clear evidence for external variability in image A on time-scales of days to years, the frequency-dependence of which supports the previous conclusion that this is predominantly due to microlensing. External variability seems to be completely absent in image B and this does not appear to be a consequence of scatter-broadening in the interstellar medium of the lensing galaxy.