Grade is king. But the queen holds the power in the kingdom and that's capital. A mine must repay capital from its margins and the primary driver behind margins is grade and metallurgical recovery of that grade. Applying a risk-weighted discount rate of in excess of 12% (yep, you read that right - mining's a risky business: what's the gold price and $AU:$US gunna be next year? How good's the geo's Mineral Resource estimate?) means that capital should be repaid within at most, five years for most operations.
Operating costs are not really relevant, because they're pretty uniform across any mine, but increase with depth. Energy's the biggest cost and is mostly driven by the mine gate cost of diesel and electricity. Incidentally, low local labour costs usually mean an expensive mine because the labour is unskilled.
KCGM and Bingham Canyon paid off their establishment capital years ago, so don't face that problem.
Most bog-standard Australian gold mines are relatively narrow fault-hosted orebodies and extend to about 120m depth, before the effects of strip ratio (ie energy cost to shift a lot of waste) makes things uneconomic.
