First, thanks for noticing the self-sabotage effect, card partly edited, but not sure if it should be buffed to be played instead of pillars as it would make them auto-includes to almost every deck.
I thought about drawpowers from other cards too. You start with 0 card, but if all of your draws cycles, like Sundials/ Scattering wind/Shard of Bravery and this could mean a stable combo win in 5-7 turns and even in those turns, you might have a Sundial down, effectively skipping minion attack. In a theoretical, optimal case where you get all the right quantas and the opponent does not disrupt you. This is the main reason why the markmatching does not give upped pillars, as it essentially would make 39 random quanta in turn 1 with Chroma mark, and generating an almost Supernova every other turn.
It's not about the starting RNG, it should be definitely worse to start with a fixed result than with something that can vary. On the upped, I'm more lenient making it have upped pillars, however, you are missing a long-term benefit. You need to put 10-13 pillars into a deck to make high-cost cards viable. But with this card, you only really need 6, which means there are significantly more actual cards in your deck than your opponent's and you can reliably set their costs to be paid by those pillars.
So the other factor is how much quanta you should be able to use per turn with them. If the upped gives upped pillars, this means you can spend 8 quanta per turn for the first 6 turns, 7 later, which changes how you construct a deck with it. IMO it's mostly a balance question as both of them can work.
