I think it's a combination. A large field in the Arctic might not be as expensive as a small field in the Arctic (in terms of overall field life costs, at least) but if there are only small fields dotted around the Arctic and in widely dispersed locations, developed by different countries and companies, then the cost of producing, say 100,000 bpd from those fields will likely be much higher than from a single large field in the Arctic, developed by one country/company/consortium.

So, you're right, location will be an important factor but the small field sizes will compound the costs.

Also, the IEA report seems to acknowledge that future discoveries will likely be smaller fields and reports that decline rates are much higher in smaller fields. This should also be a concern to them, but, somehow, they still see output rising by 70% by 2030 (if everything goes right, of course). Very strange.