Any non-zero radiation resistance will do. What matters is not the radiation resistance, but the ratio of energy radiated to energy lost to other means, such as ohmic losses in the feedline, antenna, or soil.
If the antenna system is made of ideal, lossless components, then any nonzero radiation resistance is 100% efficient. While somewhat counterintuitive it's easy to prove: the antenna must be 100% efficient because by definition there is no loss, so by the law of conservation of energy 100% of the energy must go to radiation.
There is this equation:
$$ \text{efficiency} = {R_\text{radiation} \over R_\text{radiation} + R_\text{loss}} $$
which would suggest that higher radiation resistance means higher efficiency. But it is important when using this equation to pick a single point to measure radiation resistance and loss resistance, and normalize all values to that point.
Consider that a typical antenna system consists of numerous devices which transform impedances. They may be simple transformers, but also transmission lines and even antennas themselves (which are effectively transformers between the feedpoint impedance and the impedance of free space) do transformer impedances. So a resistance of a particular value may appear to have a different value when viewed from another perspective.
For example, a folded dipole has 4x the impedance of a similar not-folded dipole, and thus 4x the radiation efficiency if your definition for radiation resistance is to measure it at the feedpoint. (There are other definitions!) But that does not mean a folded dipole is 4x as efficient, because the ohmic losses in the antenna conductors are also 4x bigger. And if this folded dipole is fed with coax it will be a poor match, and the antenna system is subject to additional SWR loss as well.