@Jam is correct — if a little long-winded — in pointing out that one of the benefits of co-transport (no need to involve evolution) is that the concentration gradient of one co-transported molecule can provide the thermodynamic energy for the transport of the other molecule against a concentration gradient. This is dealt with in a quantitative manner in Section 13.1.2 of Berg et al. and in a qualitative manner in Section 13.4.
However, there is another factor to consider. This is the maintenance of electrical neutrality. Typically it is charged ions that are transported (anions in the examples shown below, taken from Ch. 18 of Berg et al.) so that co-transport of a counter ion is needed to maintain electrical neutrality.
The diagram above also highlights another important point. You need to look at the biochemistry of the co-transporting systems, rather than just considering them in abstract. One of the sites of many co-transporters is the mitochondrial membrane, where transport is part of a co-ordinated process.
Consider for example ATP, ADP and phosphate. ATP is synthesized in the mitochondrion from ADP and phosphate, and then much of it must be transported into the cytoplasm. At the same time ADP and phosphate must enter the mitochondrion as substrates for the generation of more ATP. Active transport is clearly a non-starter here (it would use the ATP that is to be transported!) and co-transport ensures that the influx of ADP is balanced by the efflux of ATP. Phosphate influx (which is balanced by hydroxide efflux, to preserve neutrality) must be coupled to this process, although I am not aware of how this achieved (contributions welcome here).
Analogous considerations apply to NAD+ and NADH, but it is the electrons that are transported, rather than these compounds themselves, using surrogate oxidized or reduced compounds. Shuttle systems of this sort integrate the biochemical functions of the mitochondrion with those in the cytoplasm, and need to be studied individually to understand the choice of co-transporters. Section 18.5 of Berg et al. is a convenient starting point.