I'd say no. Losing Scout sucked from that standpoint since they were the second best at rankings behind 247. All the services have their flaws, including 247, and I still think the composite does the best job of smoothing everything out. Rivals and ESPN are both bad about committing limited resources to scouting, so they overrate kids who go to the camps they like and underrate kids they never see, however they don't rely on the same camps so using the composite kind of makes them even each other out.
Brayden Liebrock isn't really all that underrated. 247 has him as a low 4 star and the composite has him just outside a low 4 star ranking. I don't think that's too off. He's a 6'5, 220 lb receiving TE without elite athleticism. The fat part of the bell curve for him is being a productive player for us but not having high-ceiling, All-American type-potential, which is pretty much what a low 4 star should be.
I'm sure this next one will induce rage in some folks, but here we go. 247 is obviously low on Johnson but the composite having him at a 92 and the #201 player in the country isn't far off. He's a great runner but he doesn't have great speed, and he's definitely on the small side. He has a lot of work to do in his passing mechanics before he can be a consistently good QB. He's got a lot of potential but also a ways to go to reach it, so his composite ranking doesn't seem all that off, either. Theo Wease, now that's just crazy. Same for Arjei.
Looking at our commits, no one is egregiously ranked. Whittington (at elast as a WR) is probably a little overrated, in part because he's hit the camps hard the last two years and Jake Smith is probably underrated, in part because he doesn't do camps. The one consistent trend you can find is that guys who go to camps early and often tend to be overrated and guys who don't tend to be underrated. Late bloomers obviously tend to be underrated too. The All-American games seem to help level both this out a bit by the final rankings.
Overall though, the composite still seems to put most guys in roughly the right place (obviously, there are some big exceptions) and using it as an aggregate way to measure a class is still pretty sound.