I thought that the Snowden approach is a valid one too until I realized how wrong I was.
The first part of the bulk sample is ca. 2000t. What they did is mine ca. 1/4 ore assumed to have ca. 16g/t grade from a paying block and 3/4 waste assumed to have no gold at all. If the resource model is right ca. 4g/t should be the milled result.
So it SEEMS as if the Snowden approach yielded the correct result. But the problem is the nugget effect. The nugget effect raises the grade to 4 g/t. Nothing wrong with that if the nuggets are contained in the ore. But what if part of the nuggets are from the unpaying blocks not in the resource model? You won't catch them later on when you mine because those blocks won't get mined because they are considered waste.
The only valid approach is using a sample tower and eliminating the nugget effect from the waste areas or blocks not in the resource model. The nuggets from the paying blocks should be included of course. If Strathcona did that it means the reported 2g/t corresponds to 8g/t in the blocks from the resource model treated in the sample when there should have been 16g/t.
Your post makes no sense to me. The purpose of processing the entirety of the bulk sample is to determine whether or not the estimated grades are in line with the actual grades that are experienced in production (this process is called grade reconciliation), not to evaluate which blocks of rock should be mined except quite obliquely. If the estimation technique is demonstrated to be inadequate because it varies too much from the actual, then that info would feed back into a decision about which blocks should be mined. On the other hand, if the estimation technique is shown to be adequate (in the case where the estimates agree with the actual to a reasonable threshold), then it shows that the blocks included are probably the right blocks to include. You are putting the cart before the horse. Logically, the first thing to do is to determine whether the estimates agree with the actual. If they do, then the blocks included in the resource and reserve estimates are probably the right blocks to include.
Once again: The bulk sample is not taken from the ore blocks only. The bulk sample contains waste. If the nugget effect takes place in the ore it works for PVG. If the nugget effect is in the waste it works against PVG. Because the deposit contains the gold, but at a lower grade in more tonnes - higher costs and uneconomic.