...while paying more to produce the composite vs. what they'd pay to make an alloy. Still, I wanted to see what numbers are:

http://www.ebay.com/gds/How-Much-Gold-Is-Really-In-A-Rolex-W...

We see that the golden Rolex has most of the weight in the bracelet and I haven't seen the "full gold" bracelet on the Apple site. The body of the golden Rolex apparently has cca 2 thirds of the golden ounce of pure gold, which is at the moment not more than 700 USD. If Apple can save a half of it, it's around 350 USD saved, not counting the higher expense to make a composite which would reduce the saving.

Now if we accept Gruber's guess that the most expensive watch will be priced in the order of 10 K USD, compared to all the profits per unit, it doesn't seem to be a big percent. But if we factor the expected number of the items sold it's not something that should be ignored. Still I can imagine that they could have simply made a body with the steel "reinforcement" inside and nobody would blink.

I sure expect somebody will properly tear down the Apple Watch Edition to measure the exact amount of gold. If Archimedes managed to do this 2 millenniums ago, why wouldn't we?

Gruber's guess is still just a guess. I'm inclined to side with Marco Arment, who thinks Apple is going for more of a $5k-ish price.

I have the feeling that the gold watch portion will actually be significantly smaller portion of the 5K price (if it is 5k) Something like $999 - $1999. The rest of the price will come in the bands at 2 - 4K each.

They still have to sell the watch itself with their margins however. There's nothing stopping you from buying an aftermarket strap, and if they require you to purchase a strap with the watch -- well then it doesn't matter how they split the price up.

If the bands are forward compatible with future generations, then it becomes more attractive to spend more on the band vs the face.

Archimedes' trick only works if you know the ratio of materials inside ahead of time. Otherwise it could be e.g. some gold + some copper and have the same density as silver.

If you know it's 18k then all you have to do is weigh it to know how much gold there is. Then measuring volume can tell you the density of the rest of the alloy.

Only if the entire mass is 18k gold. Not if the entire mass is 18k gold plus other Try your approach on two cases: 0% gold, with 100% unknown alloy, and 100% watch components.

Like how foods made with "100% fruit" still have other ingredients.

They're saving on the rest of the yellow gold alloy also, which is usually silver and copper, and not that cheap either.