Although fortunately there are some minor options to alter the auto route plotting parameters, for the most part, the gps bicycle routing on the 1040 is really terrible. As someone who has bicycle toured for close to 55 years, I have some pointers that should be considered for developing an alternate algorithm that is safer as well as better for many bike tourists:
(1) The most important factor in safety on a roadway is not traffic level, - it is the presence of a paved shoulder. Frankly, it is actually safer on busier roads for the simple reason that if there is an accident, there are more people likely to stop and render aid. Remote small roads with no shoulders are downright DANGEROUS. As a long-time tourist I avoid them like the plague unless there is a compelling reason like some grand scenic interest, or it makes the route shorter. The low traffic rate, although superficially appealing, is a real danger because small roads rarely have a shoulder, are in greater disrepair, and traffic rarely slows down. The lack of traffic means some drivers are lulled into driving around like Mario Andretti half asleep, setting up bicycle riders for the greatest road dangers they could possibly experience and in a remote place where any accident would never be observed, and help might be many hours away. It is sheer irresponsibility for the current routing algorithm to be designed the way it is.
(2) As currently set up, the 1040 always picks some ridiculous circuitous route over a direct route, often completely bypassing more direct roads that have not only have wide shoulders, but often official bike lanes. The 1040 must always have outdated information, as I did update the map before discovering the map was completely ignoring perfectly good roads that had both official bike lanes as well as wide shoulders. That being said, I like it when the algorithm does have access to bike lane info as long as it can SHORTEN the route between points.
(3) I wouldn't mind the gps recommending some scenic overlook or scenic route as a POSSIBLE waypoint, but it needs to ask.
(4) Waypoint plotting is really clunky, there needs to be a faster way to line up waypoints and then have the algorithm try and route between them. Currently that would be the fastest workaround to the terrible routing system.
(5) There needs to be more emphasis on shortening distance between points. As it is, I have to turn on the "motorcycle" option to get the best route, then laboriously add waypoints at some key areas where there is an equally short, parallel option that maintains a shorter distance.
