A U.S. firm has undertaken commercial midair rocket launches for about 20 years, and the system has been studied by a number of countries.
Indeed, OSC Pegasus was flying for about 25 years now (the latest launch was with IBEX in October 2008). It teeters between marginally profitable and making a small loss, for the reasons of both high fixed and marginal costs. OSC keeps on the books and maintains a dedicated airplane for it, the "Stargazer". So, although there is no dedicated launch pad (as mentioned by the article), fixed costs are still in the millions. Further, as Elon Musk observed, Pegasus itself is a 3-stage rocket with an optional 4th liquid stage ("HAPS"), and the first stage is no less than an actual hypersonic airplane. It's not cheap to make (and is not going to be cheap to develop for Japan).
The ministry plans to develop the midair rocket-launching system at a cost of 10 billion yen to 20 billion yen, about 10 percent of that for the H-2A Greenhouse Gases Observing Satellite, named Ibuki, which was launched Friday. It also will seek to hold down launch costs to several hundred million yen.
If they attempt to clone Pegasus, the above is way optimistic. But if they are smart, they will use the concept by Air Launch (only solid fueled): a rocket which is thrown out the back of a military transport, launches vertically, and passes behind its carrier. It's a far superior system in all respects. Firstly, there's no need to maintain a dedicated airliner. Secondly, the rocket is far simpler -- no wings -- which means cheaper. And finally, there's a better growth path. We'll see if Japanese agree.