That's not the argument people are making---at least, not the people I have in mind!
Instead, the argument is this: Imagine you have a particular set of
DVC resort nights planned for the next many years. There are two different ways you can pay for those stays. One way: buy the right number of DVC points to make those stays happen, paying the entire purchase cost in cash, and then dues as you go. The other way: Take the purchase price and invest it in a low-fee index fund. Every year, you add to that fund the amount that would have gone to dues, and then you rent the DVC nights you wanted from (a) Disney or (b) from an owner at the prevailing rates. But, you will take exactly the same vacations in these hypothetical futures--the only difference is how you pay for it.
The question then is: which of those options costs more? If your hypothetical investment account has money left over then renting is a good deal. If you run out of money in your hypothetical investment account, then buying is a good deal.
It's a little more complicated, because to answer that you have to make some assumptions about how the market will behave, how rental prices will change, how Dues will change, etc. But,
MouseSavers has a spreadsheet that you can use to play with different assumptions, allowing you to come up with the answer you prefer. ;-)
I think of it a little differently. You are not only getting that extra 5%, but you are also
guaranteeing that you will get that discount on the room you want at the resort you want for the dates you want. That might matter, because your preferred room type/resort may not be part of the discounted offer for some or all of your nights, or it might be less than the expected 35%, or whatever.