Can work documented within a manual be published in a journal?

I think the comment is fair, but can be rebutted. I would prepare an answer layered on two levels:

• On the practical level, change figures xyz as much as you can. Maybe run some additional simulations, change the plots, legends, colours, etc.. This will show some willingness to the editor.
• On the more crucial aspect of whether your work can be considered "novel enough" to be published in the journal: I would argue that
  • the paper will provide a much more coherent context to the model (as only published articles can), and as such its dissemination will be facilitated;
  • the guide is subject to be changed in the future, while the published paper is a much more "stable" resource for future researchers;
  • depending on the model itself, you can argue that the paper will make it possible for researchers that don't use your code to implement the model in their own codes. If this is true, you might reword some parts of the manuscript to focus more on the model itself, which should be the paper's main selling point, than on its implementation in the code.