Holter is onto the one big idea about orchestration, and that is that innovation in orchestration always requires going beyond design specifications and received cultural associations.
The first composer to invite horns from the hunting party or trumpets and drums from the military or a trombone choir from the church tower acted transgressively, both musically and socially. The very act of integrating members of rigidly separate guilds into a single performing force must have been a bit of staggeringly diplomatic labor negotiations. And imagine the shocking effect of the first time an audience encountered such a juxtaposition: what are those hunters doing in the orchestra?
And this transgression is present at the smallest possible level of orchestration: which note (and which register, timing, dynamic, articulation) to which instrument, and in the combination of instruments. Orchestration books tend less to be theory than to be lists, catalogues, recipes; time-tested recommendations about what sounds good, effective. The musical equivalent of an etiquette book: how to be well behaved.
But great music is never well-behaved -- except, of course, when making fun of the well-behaved -- and great composers are always discovering the transgressive potential of their materials. (Quote me on this: Great composers are always rediscovering the rawness of their materials). At the same time, and particularly in the domain of timbre, a composer has to be somewhat economical with her or his inventions, as ear fatigue, preciousness, and even a certain amount of fetishing are real risks. But your mileage may vary: I will never tire of the opening Basson solo in Le Sacre in its forbiddenly high register, or the gurgling conch shells in Inlets, but for me the preciousness of the effects in Vox Balaenae didn't last longer than the month of July 1977.