Size of the intake runners does indeed matter significantly, as does length of the runners. What people tend to forget is that air does have mass. The airflow through an intake is not a nice smooth flow, in spite of the fact that magazines like to publish airflow test results.
The air column for each cylinder only flows when the intake valve is open. If the cross section of the intake runner is too small, then there is more pressure drop as the air moves, and less air actually gets into the cylinder. If the cross section is too large, then the air moves more slowly as it takes longer to accelerate the larger mass of air.
The length is important because air is also compressible and if properly tuned for length, the pressure waves can help accelerate the air column into the cylinder. That is why the SHO has primary intake runners that are longer and shorter secondary runners - to take advantage of the pressure wave to fill the cylinder better. If you have ever looked at a torque curve for the stock SHO motor, there are two torque peaks - one before the secondaries open and one after - now you know why. If you run secondaries open all the time, the low rpm torque falls significantly. If you never open the secondaries, the higher rpm torque falls significantly.
OK, so what does all that mean in terms of the op's question? Guys have done extrude hones on intakes and shown performance improvement, mostly at higher rpms. I think that is as much from the smoothening of the intake runners as from the cross section ***********, but I don't know. A good test would be needed to confirm. In any case, I would not be interested in anything that made the cross section area smaller by more than a couple of thousandths of an inch (a thin layer of paint). The thin layer of paint would not even out the casting roughness, so extrude honing or other smoothening would need to be done first, in which case I don't see the point of the paint. A layer of paint thick enough to take out the casting roughness would result in a significant reduction in cross section, which would cause me to rule it out immediately. Remember that the area of a circle is proportional to the square of the diameter, so doubling the diameter gets 4x the area.
sperold's comments about larger exhaust not necessarily being better are only partly relevant - the exhaust side has much more than atmospheric pressure to induce flow, and the reflected pressure waves are even more important, so exhaust tuning is even more challenging, imho.