How do you figure?
Say some extra weight delta_w in additional equipment is required, and say that the fuel required to complete a given flight is a function f of the plane's weight, and that the cost of the fuel is C. Further, let the lifetime maintenance cost of the new equipment be E, and the lifetime maintenance of the plane's structural components be g, also a function of its weight. As a rough approximation, the additional cost should be
E + C df/dw|_W * delta_w + dg/dw|_W * delta_w
(Note that, really, f is a function of both passenger power consumption and of weight, both of which we're varying here, but I'll buy your argument that df/dp|_P * delta_p is neglible in comparison to f(P,W))
The article asserts that one or both of the derivates in this equation are large enough that its product with delta_w is non-negligible (in fact, prohibitive). I'm guessing that df/dp is not the problem, which is the only term you really addressed.
More to the point, 240W is irrealistic for a laptop. Most laptops consume under 40W. With this better estimation, this is 1/117th of a percent.
It's more than just fuel costs.
The numbers get a bit more complicated when you figure in the weight for transformers, alternators, inverters, etc. But fuel is not a major component of this. The math really doesn't support the assertion that it's not cost-effective.