There’s a couple of issues at play which mean it doesn’t work exactly like that.

Firstly larger ocean freight don’t scale proportionately to weight linearly. So even if we assume that ebikes would mean a full doubling of the weight of a given trip it wouldn’t require double the energy and therefore emissions to do the trip. It will depend on the exact vessel but an estimate from here is for each additional 100 ton of mass to a container ship it would use an additional 0.0714 of a gallon of fuel. Its very cool physics which is largely just down to the sea doing most of the work carrying the weight itself (the same works for different reasons for rail but all other modes have much closer to linear scaling).

The other factor is that in practice the energy and emissions are the result of whole systems and trips are not always operating at ideal conditions. So its quite hard to judge what actually changes in a while system if there’s an increase in some weight of some products.

These are the reasons that additional weight in ebikes doesn’t come out to a huge increase in shipping emissions when its all worked through.

edit: paper uses imperial ton not tonne - corrected