That is nice. But can I, and everybody else in the USA successfully use Bitcoin on Black Friday? According to the National Retail Foundation, in 2016 over 101 million people went out on black friday to buy something. If Bitcoin could somehow 10x its transaction rate to even 40 transactions per second (which it can't and probably never will) it would take almost 30 days to process every order. At its current rate of 4tps, you are looking at about 292 days--during which, according to Wolfram Alpha, Venus will have made 1.3 trips around the sun.
So again, yeah, great job. Bitcoin solved some technical problem. Go team! But who gives a crap if some technical problem doesn't solve any real problem? Isn't that what we engineers exist to do? Solve real-world problems?
That shouldn't diminish all the other things that BTC does successfully.
b) Precious metals are too reliant on physical security. Someone can break into your house much easier than into i.e. your hardware wallet.
c) Besides these, precious metals are not a payment system, much less a programmable one. You can't code for example a logic that requires multiple untrusted parties to sign a transaction for it to complete.
If you've tried linking distributed Oracle databases between just a few different sites, you'd understand this has tons of failure cases and is, in fact, not a solved problem, even in the centralized world.
(Bigtable probably gets the closest these days, if you want proprietary and not globally auditable.)