Yes, that could help unclutter a schematic. Which (in my mind) helps the circuit designer communicate his original design intent. It’s the approach I have been conditioned to use, but it’s not the only reasonable approach.
An alternative I have seen is to create a section of the part that has ONLY the power connections. (The schematic symbol would be just a rectangular box with the top-level component reference designator, e.g., “U1”.) Then the individual functional sections - the six line drivers, or the four opamps, etc - would appear WITHOUT power pins. Their reference designators would identify the particular sections, e.g., “U1A”, “U1B”, etc and of course the correct pin numbers associated with that particular section. Thus, after adding the part’s “power” section, a part containing “n” functional sections ends up with “n + 1” sections in the KiCAD libraries.
With this approach, the power section of the part is typically tucked away in a corner of the schematic, shown interconnected with (only) the power sections of other parts in the design. In the extreme case, EVERY module or IC in the design becomes a multi-section part: a single opamp, for example, has one section represented by the opamp symbol, for its functional section; plus a section represented by the power-connection rectangle symbol, for the power connections.
This approach can help streamline a schematic sheet since the power connections are tucked away in a convenient corner or even a separate sheet, and not scattered over the entire diagram. On the other hand it complicates programming for EESchema as well as complicating the circuit designer’s original schematic generation. The first time a part is used, its “power” section must be placed (and probably should be connected) before its functional section(s) can be placed and used. I can visualize a circuit designer working in EESchema, going to the “Place Part” menu and calling up one NAND gate out of a quad NAND-gate package. The symbol that pops up on his (or her) monitor is NOT the NAND-gate symbol he expects, but rather the power-connection section for the 74HC00 component. “WTF??”, he exclaims loudly, waking up his wife. Not only is he aggravated because he didn’t get the NAND symbol, but she will likely NOT be in a mood to comfort him in his frustration! (Don’t ask me how I know these things.)
(Don’t real electronic engineers work at home, late at night? If you don’t think so, you need to read the “Midnight Programmers” chapter in Tracy Kidder’s “Soul of a New Machine”. See http://www.bookrags.com/studyguide-the-soul-of-a-new-machine/chapanal005.html#gsc.tab=0 )
One other question about KiCAD’s multi-section parts needs to be answered. That is, how does the symbol editor handle functional pins that are common to two or more sections? Examples might be two or more opamps that share a common “OUTPUT ENABLE” pin, an array of latches that share common LATCH and CLEAR pins, etc.
Dale