I don't know about current RC practices, but from a battery lifespan perspective, the best arrangement would be very few cells in series

Most (competitive) RC disciplines mandate the series cell count, and those that don't (drag racers) care about voltage more than capacity. I don't know whether a lightweight voltage aggregation system could be made with enough efficiency to be practical - people are dumping 3Ah packs in under 10 minutes (>6C rate), so it'd need to be able to sustain 20A with near 100% efficiency to be worth considering.

Do any RC systems use a smart power management system in the vehicle?
Do any of them regulate (voltage boost) power output, such that MORE power can be delivered to the motor near the end of the race, despite the rising internal resistance of the batteries?

Not much, with the trivial exceptions of voltage boosting for the radio circuitry itself (most important for 4 cell classes), and some newer speed controls have a voltage cutoff to prevent over-discharging the pack as a whole.

IMO the biggest hurdles for RC technology at the moment are the national and international governing bodies who seem to set the rules to protect the bigger vendors rather than to promote competition.
The biggest inefficieny in electric RC cars is the motor - ~50-60% efficiency. Until mobile phone and laptop technology recently pushed NiMH to the fore, NiCds were topping out at ~2200mAh, making the battery pack a huge factor in racing competitiveness - hence ludicrous pricing for matched packs. Recently NiMH has helped reduce that pressure, at least in the local or club levels. The air/boating RC scene have made significant moves towards brushless DC motors which boast ~85% efficencies, thus reducing (again) the pressure on the pack capacities. RC cars are lagging because brushless motors aren't sanctioned for racing.