With the advanced Gravitational Wave detectors coming online, astrophysicists hope to soon detect the signal from the merger of two neutron stars (NS) or black holes (BH). We investigate the impact of the new constraints on the birth properties of massive stars on the predicted rates for NS and BH mergers. Despite the large changes with previous assumptions (larger binary fraction, stronger preference for very tight systems), we only find an increase of less than a factor 2 (insignificant compared with evolutionary uncertainties of typically a factor 10-100). The uncertainties do not significantly affect (within a factor of 2) our predictions of double compact object merger rates. An exception is the uncertainty in IMF (variations by a factor of 6 up and down). No significant changes in the distributions of final component masses, mass ratios, chirp masses and delay times are found. We conclude that the predictions are, for practical purposes, robust against uncertainties in the initial conditions concerning binary parameters with exception of the IMF. This eliminates an important layer of the many uncertain assumptions affecting the predictions of merger detection rates with the gravitational wave detectors aLIGO/aVirgo.