I would also argue that ACH process $40T a year, while all credit and debit card networks move a collective ~$7T.
At least in the U.S, Stripe's opportunity is moving more of that $7T to ACH, not moving existing ACH to their platform.
This is poignantly addressed by @PC's comment addressing whether or not this would be available with Stripe Checkout.
ACH, for now, is generally terrible in retail situations. NO one wants to wait an additional 2-3 days (to all for funds to clear) to see their shipment be sent from Amazon. There are some niche use cases where it may work for pay-ins, for example online custom wood shops—where there's a built in delay to build the product—or other services where physical products don't need to change hands.