I agree with the other two answers, but I have a little more to add, so since you haven't accepted an answer I'll throw in my two cents.

The point of potions in the cost isn't to make things cost more, it's to make things cost differently. (The re-shuffle delay is also a good point.) If you want a potion card need to invest in a potion that won't be useful for other things. By just increasing the cost, you lose this. In games with a few Alchemy cards available, the decision to buy potions or not usually depends the variety of potion cards available. While Alchemist can be a strategy in itself and a Familiar might be necessary early on, I probably won't buy a potion to get Vineyards unless there are other potion cards I want as well.

Because of this I agree strongly with philosophyguy that, if you're proxying, just proxy the potions as well. However, if you really want another mechanic, I think adding 3 to the cost would not work well. For Familiar and Alchemist, 2 is more appropriate (equating them with Witch and Laboratory -- perfect in the last case without potions). However a University or Scrying Pool for 4 would be ridiculous in many card sets, they would need to be 5. But I think there needs to be some sort of other cost mechanic.

Perhaps require trashing a (non-copper) treasure card at the end of your buy phase. It could work reasonably well to make trashing a Silver part of the cost of the 2P and 3P cards, trashing a Gold required for Possession (and maybe Golem?), though I don't know how to extend this well to the 0P cards.

This solves the reshuffle problem adequately; you'll need to buy a Silver in your first 2 turns to get a potion card in the next two. Not being able to re-use the "potion" (can't trash a Silver more than once) is somewhat balanced by being able to use the Silver for other things first. I think it would be playable.