Yes, that's actually a very nice way to look at it. Both approval and score voting could be seen as asking two separate questions:

1. How do you rank the candidates?

2. How do you wish to allocate your voting power among these rankings?

The difference between approval and score votes is that approval requires you to allocate 100% of your voting power at a single threshold, while score allows you to distribute your voting power among various distinctions - e.g., allocate 50% of your vote to 1st versus 2nd, 25% to 2nd versus 3rd, and the remaining 25% to 3rd versus 4th.

The first question, the ranked ballot, can be answered entirely honestly, while the second question is mainly tactical in nature. It's nice how the honest part can be separated from the tactical part in this way, and it highlights the point of my article here: a standard approval or score ballot contains intrinsically tactical information for which there is no such thing as a "most honest" answer.