That Venn diagram is largely accurate, though I think ignores that while the relationship between density and service costs exists, it isn't always as simple as often presented. Generally speaking - yes, higher density areas will result in lower service costs per resident. If starting from a blank slate I'd say it's an absolute.
But - and this can be crucial - in an existing urban area the addition of density beyond a threshold can not only have diminishing returns but in fact be counterproductive (from a servicing cost perspective only - not talking about urban vibe/feel). We've reached a point in some places where land costs and the state of existing infrastructure results in exorbitant expense to scale up beyond a certain point. Neglecting this results in the situation that's already being felt in certain parts of central Toronto - schools are full, the sewer system is near peak capacity, traffic is atrocious and transit is lagging. The latter is partially being rectified but as we all know it's not cheap!
I'm not trying to do a NIMBY "no more towers!!" thing, but rather highlighting that we've dug ourselves into a bit of a hole that's becoming increasingly difficult to get out of. I don't have the answer, but I can't foresee success with any magic bullet solution. Higher taxes - whether it be at the municipal or provincial level will no doubt need to be part of it.