review a-star heuristic for different objectives

[robfitzgerald]

here are runtimes, iterations, and memory usage on a recent run of a route from DEN -> BOS. performance of the a-star heuristic seems like something to revisit.

some immediate observations:

• cost-optimal runs take longest, but, not because of tree size, but because of iterations
  • aka, lots of revisited edges
• time has the largest tree, aka, is probably getting way off target
• energy optimal, or, even 10% energy-optimal does much better with tree size, aka, a star heuristic working well here
• most queries took 2-4 minutes to run

[robfitzgerald]

@nreinicke fyi