[mapt]

Radial velocity surveys require so damn much light, and such a complex precision spectrometer that they're only used on the very largest 8m-10m class telescopes on the ground, shooting in near infrared through the most advanced adaptive optics (or even interferometric modes) in great weather, pointed at a single target for a long period of time (this is a big deal), with a focus on super-Jupiter to Jupiter class objects in tight orbits.

The next generation of 30m class telescopes will be an order of magnitude more capable for the RV method, but even then you're not really going to be able to get fast locks on Earth analogs.

The RV method is vastly superior for detecting the planets we really care about - high confidence nearby Earth analogs. The odds of a transit being in the right plane for us to observe are tiny. But if we want to run a survey like that like it really matters (let's say a Solar system catastrophe hits a thousand years from now and humanity wants interstellar diaspora), we'll be studying the nearest thousand stars with the RV method using significant numbers of 100 meter class telescopes, or perhaps big space based interferometers produced in mass quantities, for decades.

What transit studies like Kepler do is study a small patch of crowded sky (most of the stars being very distant) with the sensitivity for very rare in-plane Earth analogs, in order to get a representative sample. When I was born we couldn't say with any confidence that planets around other stars existed, post Kepler we know that they're common. We can perform these surveys even with the shoestring budgets current governments afford astronomy because even if the odds of successfully detecting a planet that does orbit a distant star are very low, we can watch a million stars at a time.