Generally the performance of a new generation of "supporting" equipment does not have a huge affect on the true performance. GDDR6 is better, but it is not the only reason the Super varient is going to be faster. Same with PCIe generations, and regular DDR3/4 RAM. PCIe Gen 3 is not saturated, so Gen 4 should not be a deciding factor (yet). DDR3 is worse than DDR4, but the performance change is seen when comparing the CPUs directly, not in the RAM alone. For both GPU and CPU, the RAM will affect the performance, but not on its own. The rest of the part needs to be able to take advantage of the upgraded RAM.
You can just compare the raw performance numbers of those two GPUs, like benchmark FPS in games, and use that to decide if you want to spend the extra $130 for the better GPU. You do not have to particularly worry about the supporting specs, like GDDR5/6, CUDA core count, or even base clock speed, because these differences will be reflected in the performance changes as a whole, so you do not have to compare those specific differences. Certain models of GPU might be clocked faster, and will perform better, but you compare those models to the same GPU in stock form. IE, not 1650 vs 1650 Super at higher clock, you would compare the base speed 1650 Super to an upgraded 1650 Super, to see how the clock speed boost changes the performance of the card over its stock clocks.
TL: DR, Dont bother with the details too much, just use benchmarks/average FPS to decide if the extra performance is worth the extra money.