I think all this talk about "devaluing" goods is a red herring. What we are in fact dealing with is
cross elasticity of demand between goods. Yes, the price of ebooks affects the profit-maximising price of hardcovers -- but the price of MMPBs also affects this (as the lower the price of the MMPB, the more people who are willing to wait for the MMPB to buy and read the book).
The degree to which prices of one good act as a constraint on the (promit maximising) price of the other is dependent on how close they are to being
perfect substitutes for each other. Given that few people seem to be indifferent between ebooks and pbooks (most either tend to move towards using their eReader exclusively, or abandoning it altogether), I would suspect that the imperfection of substitutability would be relatively high (far higher between HCs & MMPBs) -- allowing for considerable leeway for a comparatively large price gap between the two.
Either way, it is a matter for data-driven mathematical calculation of the profit-maximising combination of HC, MMPB & ebook prices, not for some panicked, sky-is-falling misconception that setting ebook prices below HC prices will "devalue" and destroy the market for HC. The latter viewpoint only makes any sense if either (i) you don't understand economics (some commenters here?) or (ii) if you lack good elasticity data and are incredibly risk-averse (publishers?).