[BETA] Improved performance (originally from #7 & #8)

[jm55DLSU]

Note that this is not intended as a deliverable but rather just an experiment

The following fixes are as follows:

1. Better search performance by using 'list(set())' as suppose to 'list(dict())'
2. Attempts to further improve path selection speed (and supposedly quality) by sorting the frontier by lowest distance first every time the algorithm searches for available paths on frontier states.

It was revealed that the fix introduces variances in the average path lengths as shown by its decimal values. As that was the case, this reveals that while the fix improved performance it did have a bit of a trade-off in terms of path selection albeit more noticeable on mazes with scarce walls rather than actual well-designed mazes.

This then was the reason why this supposed improvement is still in BETA stage as it is yet to be proven if the performance benefit is worth the minor change in path selection quality.

Before #7

During/After #7