so it's pretty clear-cut for me that Chapel isn't the strongest card in Dominion.
I don't think it is, either. I have very little experience and I hope, I understood the ranking method concerning all cards correctly, but as of now - with only a handful of people having submitted a ranked list of all the cards - all this ranking says is: "Chapel is the best card in Dominion, if a card's absolute value was only determined by how it compares to all cards costing the same." But - and I say this not, because I think I know what I'm talking about Dominion-wise, but because it seems logical - that can't be the case, since that would split up the effect of a card and its cost in a way that cannot portray the actual strength of a card accurately.
I'm not sure how you would have to go about this (in fact, all my ideas here might be older than the sun and have been put forth before, I just didn't know about the rankings then, or Dominion, for that matter), but you could either try an intuitive approach (just what the above mentioned handful have been doing presumably) to rank all cards or you could try one that tries to separate effect and cost carefully before combining them again.
First, rankings of all the cards, based solely on the value of the card effect without considering cost, would have to be submitted (super difficult and also highly intuitive). Second, you would have to determine how well you expect a card costing "X" to perform compared to a card costing "Y". If you take $2 cards as a baseline, you probably wouldn't just assume $3 cards to be "1.5 times (3/2) better" on a regular basis, nor would you even automatically assume that a $5 cost card is 2.5 times (5/2) better than an average $2 cost card. I have to admit, I'm at a loss determining how well cards are supposed to perform compared to their cost, though. Plus, you cannot just multiply effect rating by cost factor to determine a card's true value. You would also need to take into account that card effect and cost might have to be weighted differently in different "cost categories". $ 5 cards are often crucial. Their effects are so good on an absolute basis that this justifies having a lower card effect to cost ratio.
The more I write about it, the more I think I'm starting to ramble, though. Anyways I'd be curious to see everyone ranking all cards and compare the results with the list that Qvist offered. I'd contribute myself and I'd try to help out Scout. Poor guy.