There aren't particles of metallic, elemental gold unless the...

There aren't particles of metallic, elemental gold unless the grades are extremely high. The gold is all bound up in mineral compounds. That's why you can't tell visually whether the chips are gold-bearing or not (unless the are very high again). The lab assay involves a chemical process to break down the minerals first, followed by a mechanical process to extract the elemental gold.

The chemical process means that the sample is destroyed in the assay. Each new assay requires a fresh sample, so no two results will be the same. The weight/settling issue probably isn't too critical because the gold mineralisation isn't vastly different to the surrounding rock. It's not like the difference between metallic gold and quartz granules for instance. Nevertheless an experienced geologist would attempt to take a representative sample rather than a few flakes from the surface or the bottom of the container or bag (or pile).

Because the gold in the ground isn't in metallic form it won't show up on a metal detector. The minerals show up on surveys as density/gravitational/resistivity anomalies but lots of minerals do; not just gold. So there isn't much can be done without some form of drilling. I think @pangea was being sardonic with his reference to metal detecting.