To test the routing engine on the Maestro 3100, I entered several destination addresses of varying distance and compares the results to a Garmin GPS and to Google Maps. Take a look at the table below to see how the Maestro 3100 faired in test:

The results shown above are interesting: The Maestro 3100 does a decent job routing here, choosing an acceptable route.

I also tested the Maestro 3100 against the similarly priced TomTom ONE. Maestro did a significantly better job than TomTom. To my horror, the TomTom ONE chose a route that was almost 25 minutes longer all three other systems. Although the Maestro 3100 performed acceptably, it was unable to match the Garmin nuvi 660 / 680's outstanding routing engine or Google Maps' chosen route. Also, notice that the Maestro 3100 actually has us arriving with the destination address on our left, meaning we would have to circle the block or make an illegal U-turn to properly arrive with the destination on our right (although in fairness, I should point out that the Garmin nuvi 660 is almost 3 times as expensive as the Maestro 3100).

Garmin and Google differ only slightly in their chosen routes, and both get it basically correct, navigating us in such a way that we arrive with the destination on our right. Also interesting is that even though the Maestro, Garmin, and Google routed us more or less along the same route, the Garmin is considerably more aggressive with its estimated arrival time, shaving a full 10 minutes off Google's estimate, and 13 minutes shorter than Magellan's estimate. Having personally made this trip many times, the Garmin chose the route I would consider "best", although Garmin's estimated drive time is too optimistic -- the actual drive time tends to be closest to Magellan's estimate of 1 hour 40 minutes.

It appears Maestro's routing engine is identical to the RoadMate 2000 / 2200T, CrossoverGPS, and Maestro 4040, as all three Magellan systems produced identical routes in my testing.

As a second test, I entered the same starting address in Connecticut, and as a destination address I entered: 17 N. 6th Street, New Bedford, MA. Take a look at the results in the table below:

On this test, the Garmin nuvi 660 and Google Maps produced identical results (although Google is slightly more verbose in its instructions). All three chose identical routes once we were on a major highway. Maestro chooses a slightly different route getting us from the starting address in Connecticut to the I-84 highway onramp. Again, Garmin's routing engine is more aggressive with its estimated travel time -- Garmin's estimate is 21 minutes shorter than Google Maps', despite the fact that they both have chosen the exact same route. Over time, I have found that you need to consistently drive faster than the posted speed limit in order to meet Garmin's estimated arrival times.

The routing engine on the Maestro 3100 produced acceptable results, soundly beating TomTom and Alpine. However, throughout my testing I noticed that if I strayed off course, although the Maestro would almost instantaneously "re-calculate" the route, it would keep trying to route me back to the original path instead of accepting the change and re-routing on the new road. In some cases this added 10 miles to the trip instead of just re-calculating the route going forward. In a side-by-side comparison along the same route, Magellan's routing engine had me doubling back (adding 15 minutes to the trip) while Garmin's nuvi GPS simply re-routed me along the road I was currently driving on (although the Garmin did make me wait a nail-biting 30 seconds before it finally re-calculated).

In my opinion, Magellan's routing engine isn't quite as good as Garmin's, although it does come in close 2nd place.