My solutions are not so different philosophically. I tend to like having a single driver carry the majority of information, and to cover the entire vocal range.

In my smaller systems, the main driver covers bass and vocals, and the tweeter only covers the frequencies well above the vocal overtones. It's like a single driver with a supertweeter.

In my larger systems, the midrange is run without an electrical crossover. It's like the main driver in my smaller systems but there is a sub added below them. The woofer is sent only information below that which the midrange is capable of reproducing. Likewise, the tweeter is crossed over where the midrange is out of gas, and both response and directivity are matched. So this is like a single driver with a built-in sub and a supertweeter and where the directivity of all subsystems is matched.