I had read that same claim elsewhere, before the publication of the book (I think Chad Yakobson was the first to describe the over-production of phenols by brett exposed to too much dark malt). I've (somewhat blindly) follow this advice, and generally replace about half of the dark malts with their dehusked equivalents. To date that method has worked very well - I've not once developed phenolic flavours at a level which was unpleasant. I've had beers that were upto 20% dark malts (e.g., SRM's in the 50-70 range) brewed this way that were delicious - i.e. beers containing 10% "normal" dark malts + 10% debittered. As such, I think your approach should be fine.
That said, I've also not experimented much in this area. Likewise, phenol production varies greatly between different strains of brett, so your 14% may be good with some strains but not with others. To be cautious it may be worth avoiding the more phenolic strains on your first attempt.
On a non-brett note, I've found that I enjoy the character of the 50/50 normal/debittered dark malt mix far more than 100% normal malt version - even when brewing conventional porters and stouts. You can really bring out the coffee and chocolate notes without excessive astringency or dryness that often arises when you push these aspects of your malt bill. You may want to try replacing some of the normal malts with debittered just to experience the difference it makes in your beer - with or without brett.
Bryan
