Latimer Oxidation States
In The Oxidation States of the Elements and their Potentials in Aqueous Solutions, Prentice-Hall, New York (1938) Wendell M. Latimer (1893-1955, UC Berkeley) scans through the elements and proposes oxidation states for each of them. Though it is not very explicitly stated, for this purpose bonds are considered ionic.
About as close as he comes to subtlety is for Hydrogen (p. 27) where he writes,
"Between these two classes [acids and "true" hydrides like NaH and CaH2] are many compounds, especially with carbon, in which two electrons are shared between the hydrogen and the other atom (covalence) and in which it is therefore ambiguous to talk of the oxidation state of the hydrogen. However, it is customary to refer to the oxidation state of hydrogen in all compounds except the true hydrides as +1 without raising the question of the distribution of the electrons in the bond."
Similarly for Carbon (p. 118) :
"we have thousands of compounds in which it is futile to attempt a classification of the carbon atoms in terms of the oxidation state. Of course, one may always write a half-reaction relating a given compound to the element, as, for example, for benzoic acid,
C6H5COOH = 7C + 2H+ +2H2O + 2e-.
But as such a half-reaction has no chemical significance, this classification of carbon in benzoic acid as having an average oxidation number of -2/7 is not useful."
[5 @ -1 ; 1 @ +3]
Even in this kind of ambiguous situation organic chemists find it useful to assign formal atomic oxidation states to individual atoms (and to molecules) of starting materials and products in order to help select reagents that might help achieve such a transformation.