Yes, based on dimensional weight unless the actual weight is greater.
Dimensional Weight is used if package exceeds three cubic feet (5,184 cubic inches). E.g A 3 cu ft package has billable dimensional weight of 5184 / 166 = 31 lbs regardless of actual weight
See this for details
http://images.fedex.com/ca_english/customersupport/faq/DimensionalWeight.pdfYou can test package sizes here.
https://www.fedex.com/ratefinder/homeThe calculation will indicate whether it used dimensional weight or not on the last page
Dimensional Weight of your package(s) was used to calculate your rate.
Yes, it does seem odd that the shipping plugins don't use dimensions, but that's because without true "packing", there is no way to determine the actual box sizes. I.e. totaling the H W L of shopping cart items the way it does now would not yield true packed box sizes.
A "true" Packer plugin would be set up with various box sizes, and would determine an optimal 3D bin pack. It's a computationally expensive process, and difficult to optimize performance characteristics (typically, heuristics are used). I don't think you will find any eCommerce system that supports it, except maybe one that costs $20K+.
Ctrottier is working on some enhancements (a fork version of nopC, until it's adopted), that will call a 3rd party 3D Bin packer API - solvingmaze.com. It does, though, have some limitations on # of box types, and #of packed boxes. Also, as indicated above it won't be prod ready for a while.
The Shipping Director "packer" does "fabricate a box" with dimensions - i.e. it creates a single dummy product variant / shopping cart item with the box dimensions, but unfortunately the Fedex plugin does not use them. I'll look at the fedex plugin, and see what's involved in fixing that (single item uses those dimensions, else use cube root as per above). The Shipping Director "packer" is however limited to a single box type (HxWxL). However, it will after packing "shrink" the dimensions on packages that are not "full" (e.g. so that if only n% of box is filled, then it assumes it will actually be packed in a smaller box using cubic root calc).
(P.S. I will post some of above to the forum thread, for benefit of others.)