Glad to see that my anecdotal experience with Google Maps and Waze turned out to be correct[0]. I've noticed, consistently, that Waze underestimates the drive time that it quotes and I've also had the sneaking suspicion that the routes weren't
perfect (though I had no expectation that they would be; there's a lot that can go wrong on a 40-minute without traffic commute).
I think one of the reasons for Waze under-estimating route times has to do with it not properly correlating traffic speeds provided by its many users against the route selection. For instance, my route involves changing from a local highway that ends to an interstate where the left two lanes are the lanes you have to be in to get where I'm going. These lanes are often moving at around 5 MPH on average whereas the other side moves at around 60 MPH. Waze will indicate that the freeway is moving at an average of 40 MPH and will leave me on that road, rather than taking an easy, single-light traffic-free shorter route around the problem. I've never spent more than 5 minutes navigating the route-around, but I've had times where staying on has taken over 20 minutes (and 5 minutes would be fast most days). It rarely suggests the route-around (despite Waze having introduced it to me). It isn't intelligent enough to know that if I'm taking that exit, I'm the 5 MPH traffic, not the 60MPH traffic. It's either not taking into account that my route would result in slower traffic than average due to lane position, or it is and the other Wazers it's basing its estimates on are less conservative drivers than me -- I tend to get in the lane I need to be in early rather than flying around traffic and pushing my way in. It's got to be a really hard problem to solve, in general. Padding route estimates or estimating toward the worst-case scenario would probably be a good idea.
That said, Waze is the only app that has actually hooked me into using it on every trip, every time. The Waze app gives me speed limits and warns me when I am exceeding them by a pre-set amount[1]. It's also pot-hole season where I live and due to the wide user base of Waze and its hazard reporting, I get a warning (well, lately, hundreds of warnings) when there's one up ahead. By far the biggest benefit, though, is the police warning. I don't intentionally speed; I generally subscribe to the rule of "keep up with traffic" while keeping to the right-most lanes[2]. That said, keeping up with traffic has gotten me pulled over once (no ticket due to a decade of not having a ticket and my only exceeding the speed limit by barely 10 MPH).
In the last year, Waze has beaten me spotting the hidden police car every single time. I went on a 3-hour drive with a close friend who has one of those supposedly "covers everything" radar detectors and on that trip Waze spotted four more cops than his radar detector[3] did and had one partial false positive (he had moved ahead about a mile, so the app showed two cops instead of one) whereas his detector went off so many times in the city that he had to silence it. The avoidance of traffic fines and added insurance cost, alone, makes it worth it[4].
I'd say, for me, I started using the Waze app on the promise that it would get me around some of the unexpected traffic difficulties in my daily commute, but the reality is that after a few months, I know the areas involving my commute far better than Waze or Google. Both navigation apps have made me a slightly more intelligent driver. I'd often skip-off the freeway for surface streets when things got slow and was surprised when navigation apps wouldn't suggest to me to do that. It turns out that the time spent at traffic lights in traffic well exceed a slow freeway. It feels faster, sometimes, to exit the wall-to-wall 15 MPH slog, but you trade it for periods of 40 MPH traffic mixed with longer periods of just sitting idle.
But the bigger eye-opener that Waze seemed to do, which Google Maps didn't -- when there were lane closures, it'd have me enter the freeway at the first on-ramp before the lanes re-opened, rather than at some point after the lanes re-opened. And every time I ignored it, I ended up in some creative form of traffic hell. After thinking it through, I realized that the first on-ramp after traffic re-opens is often flooded with everyone else who's trying to avoid the construction. The last on-ramp just before the lanes re-open is not similarly flooded and yet traffic on the freeway is moving at about 60% of the speed-limit at that point and only getting faster as it reaches the re-opening. Since the last merge bottle-neck is that on-ramp and it's not clobbered with diversion traffic, it's a very good workaround.
[0] I don't own Apple products, so these are the two "big choices" on Android.
[1] I bumped it up from warning at the actual speed limit because driving the speed limit on some of the roads I travel is more likely to cause death-by-someone-elses-road-rage.
[2] Where I live, two-lane roads require drivers to remain in the right lane and use the left lane for passing and this rule is enforced ... just about never. Multi-lane roads are fair game with most remaining in the middle one or two lanes for the entire duration. I'm not even sure what the law is for these but if I were to judge based on how 90% of people drive, I'd say the law is to never be in the right lane, and that the left lane is for people traveling more than 20 MPH over the speed limit. We have no HOV lanes.
[3] I can't write that without thinking Radar Detector Detector Detector Detector https://rationalconspiracy.com/2016/06/02/radar-detector-det...
[4] "Well, you can avoid that by just not breaking the law." Except that traffic laws are among the few laws that are easily broken unintentionally.