My idea of how to evaluate the candidate is completely orthogonal to what you are asking.
1) What counts towards evaluating your research potential is what you proved and what tools you developed. You can put 15 theorems into one paper, and the theorem count will still be 15. You can repeat the same idea in 20 different variations in 20 papers and the idea count will still be 1. You can have 3 tricks and publish 7 papers with all possible non-empty combinations of them and the trick count will still be 3.
2) What is derived from how you split and group your results for publication is your maturity as a writer. At the graduate or postdoc level, as far as I can tell, almost nobody cares about this aspect because it is commonly assumed that the choices here are made not as much by you as by your mentors and advisers (especially if the publications are either joint with them, or just thank them for helpful advice anywhere in the text).
3) Ideally, a single paper should contain a single statement. This statement may be simple or complicated, long or short, a startling novelty or a small twist of a routine, etc., but it should be a statement that can be understood and digested as a single block like a sentence in a book. Of course, it is not always possible, but still this is what (in my eyes) determines where to put a comma and where a full stop when writing. The other considerations are far less relevant because you write not for the members of hiring committees, but for unknown people for most of whom you exist merely as a combination of the ideas you share and who do not care in the slightest about your personal status or reputation.
4) With all that said, if you want to land a good job, you need to show up on radars. So, write sparingly and concisely, but talk profusely. Don't hesitate to go to conferences, to meet with people, and to use any other opportunity to get acquaintances. Quite often "I see John is applying..."; "Yeah, Peter told me he would..."; "They also consider him at..." can secure you a position better than "Look, this theorem is just brilliant!"; "Theorem by whom, you said?"; "I cannot tell much because it is so far from my field, sorry...". The second approach works too, but you need to be really good to just throw things into the wind and see how they soar higher and higher. Most of us, poor mortals, need to hold the strings of our kites firmly and pay attention to their tension :-).