We characterize the intermixing of network-modifying Na+/Ca2+ ions around the silicate (Q(Si)(n)) and phosphate (Q(P)(n)) tetrahedra in a series of 16 Na2O-CaO-SiO2-P2O5 glasses, whose P content and silicate network connectivity were varied independently. The set includes both bioactive and bioinactive compositions and also encompasses two soda-lime-silicate members devoid of P, as well as two CaO-SiO2 glasses and one Na2O-SiO2-P2O5 glass. The various Si/P <-> Na/Ca contacts were probed by molecular dynamics (MD) simulations together with heteronuclear magic-angle-spinning (MAS) nuclear magnetic resonance (NMR) experimentation utilizing Na-23{P-31} and Na-23{Si-29} REDOR, as well as P-31{Na-23} and Si-29{Na-23} REAPDOR. We introduce an approach for quantifying the extent of Na+/Ca2+ ordering around a given Q(P)(n) or Q(Si)(n) group, encoded by the preference factor 0 <= P-M <= 1 conveying the relative weights of a random cation intermixing (P-M = 0) and complete preference/ordering (P-M = 1) for one of the species M, which represents either Na+ or Ca2+. The MD-derived preference factors reveal phosphate and silicate species surrounded by Na+/Ca2+ ions intermixed nearly randomly (P-M less than or similar to 0.15), except for the Q(Si)(4) and Q(Si)(1), groups, which manifest more significant cation ordering with preference for Na+ and Ca2+, respectively. The overall weak preferences are essentially independent of the Si and P contents of the glass, whereas P-M primarily correlates with the total amount of network modifiers: as the latter is increased, the Na/Ca distribution around the {Q(P)(0), Q(Si)(1), Q(Si)(2)} groups with preference for Ca2+ tend to randomize (i.e., P-Ca decreases), while the P-Na-values grow slightly for the {Q(P)(1), Q(Si)(3), Q(Si)(4)} species already preferring coordination of Na. The set of experimental preference factors {P-Ca} for the orthophosphate (Q(P)(0)) groups extracted from P-31{Na-23} REAPDOR NMR-derived M-2(P-Na) dipolar second moments agrees well with the MD-generated counterparts. Our results on the Na/Ca intermixing in soda-lime-silicate glasses are discussed in relation to previous reports, highlighting the dependence of the conclusion on the approach to data evaluation.