There's always a tradeoff between R&D costs and operational costs. From what I gathered from the AW&ST article, a Mach 7 scramjet first stage could be developed relatively quickly using existing jet engines. The jet engines would get the vehicle up to about Mach 2.5 where ramjets would take over. Existing jets would do the job. It'd be interesting to see if the hydrocarbon fueled scramjets that the Air Force is researching would be a good fit for this design. Being able to use hydrocarbon fuels would simplify the fuel system and allow for a smaller vehicle due to higher energy density.

From what the article implied, if you want a faster scramjet, you'll also have to go a lot faster to ignite them (Mach 4+). That means you'd probably need to develop new jet engines, an expensive process. You'll also have greater thermal problems to deal with. All of that would push up the R&D costs considerably. Even if the faster speed might result in lower operational costs, the combined R&D amortization costs with the operational costs might be too expensive.

This may be another case where the perfect is the enemy of the good. While a Mach 10 first stage might be more "perfect" than Mach 7, the R&D costs of developing the faster scramjet first stage could be so high that the project is a non-starter. A 20,000 pound payload to LEO booster with lower launch costs and high reliability is worth consideration, IMO.