They are passive yes, so what we mean by gain is the comparison of one type of antenna with respect to another reference antenna. For example the gain of a half wave dipole is 2.15 dBi (directive gain relative to an isotropic radiator). Compared to other types of antennas, the half wave dipole might have less gain. It's all relative.
This. You could take a coaxial cable and stretch the shield and conductor out in opposite directions, and you will have an antenna, in fact a dipole. It will have more RF signal gain in the directions at right angles to the conductors, and much less gain in the direction along the conductors (the overall energy delivered in all cases is of course less than 100%). It will in fact have a wave impedance of about 72 ohms, which means it won't transfer power to the E and B fields as efficiently as something like a folded dipole, even if you can deliver 100% of the power to the element itself by matching its electrical impedance. In practice matching the signal impedance is more of a concern than matching the free space impedance, though both will improve the antenna performance. Optimization of the antenna design for one over the other depends on many competing factors.
The total power radiated (or received) by a certain type of element can also be "forced" along a certain direction more efficiently than others by shaping or changing the configuration of the conductor(s). A parabolic reflector behind the active elements, for example, can make the effective signal many dB higher in that one direction (I've worked with radar antennas having as much as 30 dB higher pattern efficiency along one axis). Passive elements in front of and behind a dipole can change the radiation pattern in interesting and very useful ways too. This changes the free space impedance, but the improvement from directing the energy often overwhelms any potential losses from that.
Also, just for completeness, the impedance of the medium you're radiating into is actually proportional to (square root of) the ratio of the magnetic permeability to the electric permittivity of the medium. For free space, that works out to 377 ohms. If you take the product of those two numbers instead, find the square root and then invert it, you will get a number equal to the speed of light in that same medium (or the speed of radio propagation). This was Maxwell's contribution to the laws of physics, wherein we understand that light and radio waves are the same thing.