That's not quite the right read - the previous work (I'm not even sure if it was that lab) was to engineer e. coli to make psilocybin, but e. coli was a bad factory because it still needed a very expensive to make/obtain chemical added as a feedstock.
So instead, they put the same genetic material into yeast which already had the necessary metabolic equipment to make the entire compound "de novo" (i.e. from scratch - just add sugar and yeast nutrients and it pumps out psilocybin). The downside is that the yeast's metabolic pathways chew up about half the product, so there is still work to do on making it more efficiently, but at yields of 600mg/L it's not exactly a bad start. (For reference, depression researchers tested the compound at dose of 30 mg for a 155lb man and it yielded psychoactive effects.)
So instead, they put the same genetic material into yeast which already had the necessary metabolic equipment to make the entire compound "de novo" (i.e. from scratch - just add sugar and yeast nutrients and it pumps out psilocybin). The downside is that the yeast's metabolic pathways chew up about half the product, so there is still work to do on making it more efficiently, but at yields of 600mg/L it's not exactly a bad start. (For reference, depression researchers tested the compound at dose of 30 mg for a 155lb man and it yielded psychoactive effects.)