The Short answer is: no, you cannot alter the .qdif text file without needing to relinearize. The simplest thing is to just alter the image to match what you want to print.
The long answer is: Lowering the ink limit will make that shade lighter, and lowering the grey value will make it intersect the next darker ink at a different point. Doing that will unpredictably change the shape of the overall greyscale, but it will not really have much of an effect on the overall density. The main point is that anytime you touch the ink limits and/or grey values you will need to relinearize your curve—there is no getting around that. This is one case of being a little lazy (and reckless) with the ink limits could introduce a whole lot of other problems down the road, and maybe ruin what was possible a good/functional profile. This is why people have problems linearizing their curves when creating their own profiles and then can't easily track down what exactly is causing it. If you don't have the correct grey values (or just calculate inaccurate cross over points) then you will have more trouble later in the process.
I would suggest that you use the eye dropper tool and the info pallet in photoshop to see exactly what grey values are blowing out in the highlights. Then print a 51 step tablet included with QTR (with the same paper and profile or just include it as a layer in the print in question). Then you will see if that same value is also blowing out or printing too hot on the step wedge as well. If there is no obvious problem with the step wedge then just use that as a guide/target tone for adjusting the image for print in photoshop. If the stepwedge looks different than your display then maybe your display has the wrong luminance/gamma/white point for the print viewing environment.
I tend to keep a printed stepwedge with that paper and profile right by the workstation to see if my 3-5%k will print as light/dark as it is perceived on the screen.
Richard Boutwell