Tentacles: The Zerg "tentacles" could be more than just additional macro. It could also be a defensive strategy against attacks such as a storm drop. The tentacles would be immune to storm, and thus a Zerg player with a tentacle-mining base would not have to worry so much about abrupt stops in harvesting. The necessity for making tentacles should be properly balanced so that a scouting Zerg player would weigh-in the decision about making tentacles early if he sees templar tech.
The tentacles themselves should cost minerals, either by permanently using a drone to morph, or being built from the hatchery. The added macro would come from the fact that there would be 8-10 mineral patches, and you would not be able to immediately create tentacles for every one. There would also be no horrible penalty for not making tentacles; after all, once the minerals run out, the tentacles are lost value. In general, tentacles would be the way to go if you had the APM to rush back to base from time-to-time to add another tentacle.
Tentacles should be destroyable so as to add more options to an invading army. Do I have enough time to take down the whole hatchery, or should I take out as many tentacles while I still have troops in Zerg's base?
To sum up tentacles: tentacles would provide certain defensive advantages (vs. templar, etc.), and boost mineral production for the cost of minerals and APM.
Protoss Mineral Harvest: While I liked the idea of this, I think it needs more work done to further diversify it from the Zerg Tentacle idea. One idea I thought of would be a mini-warpgate (I'll dub them a name like "mineral-fold" or "minefold") buildable by probes upon each mineral patch. Instead of bringing the minerals to the nexus, the probe would actually blink off the screen - transferring minerals to some other Protoss nexus across the galaxy - and blink back to immediately begin harvesting again. The result would be quicker mineral harvesting, but at the cost of a minefold for each mineral patch. The macro component would come in to play because a player couldn't simply spend all their initial minerals on minefolds at once. The construction would have to be dispersed over a period of time, dependent on the developing military conflict. There would be natural bonuses for these minefolds: if probes are off the screen half the time, speed/targeting-dependent economy harasses would be less optimal. Additionally, if the Protoss Nexus is destroyed, probes would still be able to mine any patches with minefolds! (certainly mineral-folds would not be buildable until a nexus is first built.) These benefits would have to be outweighed by the cost or time required to build a minefold. For example, when a probe starts construction of a minefold, that mineral patch could be un-mine-able until completion. The Protoss simply wouldn't want to convert all mineral patches to minefolds at once, since this would result in a deadly delay in harvesting.
To sum up minefolds: Probes build "minefolds" on individual mineral patches to decrease harvest transport time and thus increase production. "Minefolds" grant additional defensive features. The cost is minerals/APM.
That was my (more than) two cents, glad to see some exciting ideas about increasing macro in this tread!