>>14279809Obviously you want more Isp than a cold gas thruster. However, the amount of Isp you actually NEED is only the amount that allows for easy-to-achieve structural mass ratios that still produce a delta V budget of over 9km/s of delta V on a TSTO.
I know that sounds like a lot of gobbledygook but really it just means "If your rocket engine has enough efficiency that you can get ~5 km/s out of a stage that is 80% propellant or less, it's good enough for Earth launch". Of course there are other factors not considered in Isp, such as propellant density, which is why even though hydrolox engines can achieve >360 Isp at sea level and >450 Isp in vacuum they're still pretty shit for reusable launch vehicles (due to having much larger propellant tanks for the same mass: nuclear thermal rockets are even worse, several times over, despite being ~twice as efficient).
The Falcon 9 is a perfect example. Less than 300 Isp at sea level, barely over 300 Isp in vacuum for the first stage, and about 340 Isp for the upper stage. 120 seconds lower than the RD-0146. However, since Isp is not that important for launch vehicles, the Falcon 9 gets >15,000 kg to LEO per reusable launch while being cheap. The way to design launch vehicles is to pick technology that is cheapest to mass produce while being good enough to physically accomplish the delta V requirements, then scale it up until you are getting the payload masses you want to see.
The cases in which Isp is actually important to chase after are ones where the single-stage delta V is very high, due to transfer delta V or lack of available propellant refilling options or whatever else. One example is a Lunar orbital shuttle that goes from the surface to NRHO and back in a single stage, carrying significant payload masses in both directions. That kind of vehicle only needs to have a total delta V of around 5.5 km/s, however increasing the stage Isp has a direct impact on greatly increasing payload mass fraction.
