Matt has created a wonderful new mechanic in D'Up (kudos!), and I, for one, want to design cards for it that allow these cards to be playable in constructed, because I want them to be a showcase for Matt's creation! I think it deserves more than unplayably overcosted rehashes of decent-to-good cards, so I've been aggressive in my costing to avoid that, and to try to allow even the staples to be eminently playable. As an aside, this is an amazing opportunity for the TMD Design Team (TM) to take a rich mechanic (richer than he thought, one suspects), and mine the hell out of it, all in the name of seeing just what we can achieve. If ever a mechanic has been created here that could be the cornerstone of a TMD set (or Block!), this is it.
A key point of Magic resource theory is that as cards get more expensive, the effects should increase faster than a linear mana comparison would suggest (something everyone here knows already in terms of deck design, but I don't think we take it into account enough in costing effects). In the case of Sorceries, 4 mana is the flashpoint at which cards should either have substantial utility (Creeping Mold) or substantial effects (Wrath), and in the case of Instants, I'd say 5 mana. This makes costing D'Up difficult at times, because undercosting is bad, but overcosting something in the name of utility is very easy, mainly because
utility becomes largely irrelevant above a certain mana cost.
My basic guide, therefore, is to look at the first three iterations and compare them to existing cards (much as you've done with your card-draw example), and if they stack up well up to that point, it's not relevant after that.
Anyways, onto the cards themselves...
Everything is better than Demystify – it was created as a bare-bones card solely for the purpose of putting in Eighth Edition. Wax/Wane, Quiet Purity (Arcane), Erase (RFG), Aura Blast (cantrip), Monk Realist (1/1), Kami of Ancient Law (2/2, sac.), Peace and Quiet (2 enchantments) and Serene Offering (gain life = to CMC) are all better than Demystify in their own way, so thatt's not saying much. The effect seems to cost White about 1 mana (i.e. a bit less than W), going by most of the cards printed, but I can see costing this variant at 1W (done). More than that, though, and it would be terrible, so nobody would use it (1W, 1WW, 1WWW etc. looks fine to me).
Rain of Salt is the direct double effect of Stone Rain. However, nobody ever played Rain of Salt, because it costs 6, and 6 mana sorceries should do substantially more than just destroy two lands. Hell, for 7 mana, Red gets an instant, uncounterable, cantrip Armageddon. In my books, a structure of 3R, 4RR, 5RRR (etc.) is fine, because it will never be broken at any of those costs. While I can understand why you left Stone Rain off, my Stone Rain variant is just the basic D'Up effect (extra mana = extra copies). There's plenty of room for other variants to exist with different cost structures, and different D'Up costs (e.g. sacrifice a land), something that holds true for a lot of these effects.
For the other cards:
Shatter – 2R, 2RR, 2RRR (looks fine to me)
Naturalize – 2G, 3GG, 4GGG (again, looks absolutely fine to me)
Dark Banishing – 1BB, 2BBB, 3BBBB (I have to say, as an uncommon or rare card, I just can't see this tearing up Limited or Constructed. It's a good, solid card, but it's not broken at any of those costs – Reckless Spite killed two creatures and cost 1BB and 5 life)
Lifegain – 1W, 2WW, 3WWW (looks fine to me, although it might actually be slightly overpowered or undercosted)
As a side note, using a Blue draw spell as a template for costs is flawed, because card-drawing is always potentially playable as long as it's not hideously overcosted. It's not conditional on anything other than available mana, and drawing additional cards is almost always good
.
