With precise, programmable control over biochemistry, we can make almost any organic carbon based molecule from almost any other carbon source- but obviously not things like metals. However, I posit we will be able to make things with drastically superior performance that fills all of the same use cases. Consider for example that Dyneema - which is just simple straight saturated carbon chains- is already 15x the strength of steel on a weight basis. I'm talking about being able to predict the properties of a molecule ahead of time, and then make something with exactly the properties we want.

It would be quite shortsighted to make a more environmentally friendly way of making the exact same stuff when those things were limited by constraints that no longer apply and we have the potential for drastically superior materials (stronger, more durable, lower toxicity, more recyclable, etc.) for a specific problem- but it depends on what specific problem you are addressing.

Basically, replace most electricity with photonics, the rest with ionics. We have efficient ion-flow based computing and flying machines, don't we? (Birds & brains)

As for how to revamp the rest of the "irreplaceable" material culture not based on photosynthesis: What's in it for me to talk to you about this :)? How many years further have you seen than Grothendieck, since Fuller was not so visionary after all? (Including, how to fund actionable, scalable stuff, that's 30 years give or take 5)

(Note that there are siliceous, if not entirely silicon-based, lifeforms on earth. Diatoms, molluscs, etc, perhaps a significant amount of our low-end chips already come from them, through seasand? :)