But it's not. You can do an RCT of the new treatment vs the old treatment. You won't get a direct measure of its absolute efficacy but you will know if it's superior/ non-inferior to the best known thing. And then you can use observational techniques to estimate the absolute values. That's exactly what you would do if you wanted to develop, say, a new flu vaccine that you thought would outperform current vaccines. You get the most important information: whether or not we should switch to the new one.

If you have a new vaccine for a disease for which there is no existing vaccine you do a standard placebo controlled RCT which gives you a direct, high quality measurement of efficacy and side effects.