I thought the bit where knitting hadn't been invented yet was pretty illuminating on the dodecahedron hypothesis.
Knitting predates the Roman empire by a comfortable margin.
The link, and wikipedia, disagree with you.
Do you have any evidence for that? Note that knitting is a specific thing, we're not talking about all woven textiles generally.
The Romans were buying knitted and spun textiles from the north-west of Scotland before they even set foot in the south of England. It's pretty well-documented.