The observed Oosterhoff period-metallicity effect of d log P/d([Fe/H])=-0.122 at the blue fundamental edge (BFE) of the instability strip can be reconciled with predictions from the canonical HB models if the observed period shift is not only a luminosity shift but a temperature shift as well. If so, predictions from models should not be made at constant temperature but at a temperature that varies with [Fe/H]. The sense is that low metallicity RR Lyraes have lower temperatures than those of higher metallicity by DELTA log T(e)=0.012+/-0.006 per dex in [Fe/H], determined from observations of field RR Lyraes. RR Lyrae absolute magnitudes, calibrated from the pulsation equation using the data on periods, masses, and temperatures at the BFE are M(bol)=0.36([Fe/H])+1.04, and M(v) = 0.30([Fe/H])+0.94. These luminosities are brighter by 0.2 mag at [Fe/H]=-1 and 0.3 mag at [Fe/H] = - 2 than those given by current Baade-Wesselink (B-W) data. Confirmations of the need for brighter luminosities come from (a) comparison of the mass distributions of white dwarf and PN central stars with RR Lyrae masses computed from the pulsation equation using B-W luminosities, (b) comparison of the globular cluster luminosity functions in the Galaxy and in M31, (c) theoretical predictions of the RR Lyrae luminosity level from current HB models with and without evolution, (d) comparison of Cepheid and RR Lyrae distances to galaxies in the Local Group where both types of variables are observed, (e) a pulsation analysis by Simon and Clement based on Fourier decomposition of light curve shapes, correlated with basic parameters including luminosities, and (f) Eggen's calibration of RR Lyrae luminosities from intermediate band photometry over the light cycles of approximately 50 field variables.