My POET: There are other engineers who would say all roads lead to 7nm and today there are engineers in a foundry who would say all roads lead to 22nm FD-SOI.

The 7 nm engineers might be right and Geoff Taylor even might agree. Why shouldn't POET's ways lead to 7 nm, too? However, when this happens, silicon will be stuck there for years already (okay, perhaps at 5 nm).

However, manufacturing technologies for 7 nm might have matured by then, making the production of POET chips cheaper than for silicon today – except there's no mass production of silicon chips at 7 nm yet, and IBM can neither say when or if this will be possible nor how much it will cost.

Regarding FD-SOI, why not FD-GAOI one day? FD-SOI is the presence, POET is the future.