nah the current model you only need 1 single SP more than the current amount of SP voting for the highest ranking witness, and you can elect not just the top 20 but the top 30 witnesses. Changing to 1 Sp = 1 vote makes a huge difference in two ways. #1 the multiplier effect mentioned before - a hostile attacker who had as much SP as the current top witness vote getter would only be able to elect ONE witness into the top 20, instead of replace ALL.
BUT more importantly, the way the hard fork rules require 17 witnesses to pass a hardfork and conversely only 4 to block a hardfork plays into 1 sp = 1 vote very nicely. A hostile takeover has to split its SP among 17 witnesses, while a community defending against a hardfork can concentrate all its SP among just 4 witnesses. That alone means the hostile party needs 4 x more SP.
Another way to visualize what a giant difference there is between the two systems is to imagine the case of voting for only one witness vs. voting for 30 witnesses. Under current system if you only vote for one witness you are in essence only using 1/30 of your total "witness voting power" (for lack of a better term). And if you vote for 30 witnesses, each one gets as much SP as the one would get if you only vote for one witness.
Now compare that to one SP = one vote. If you only vote for one witness (spending all your SP votes on that one witness), you are using 100% of your witness voting power. It is all just concentrated on that one witness. And that one witness vote would be 30 x as powerful as the votes you could cast for 30 witnesses.