With 3.0 we've worked hard to pull in the v1 API. Both the v1 write and query endpoints are supported in v3. There may some gaps here and there, but our goal is to make it so that all the things that worked with v1 will work with v3. The only two exceptions would be the subscription API that Kapacitor used and Continuous Queries. We warned people off from using both of those features in v1 as they didn't work well under load.
With v3, I prefer to think of it as us doubling down on core database performance and functionality. With v2 we tried to create this whole development platform. V3 brings our focus back to the core database, which I think will yield better results for everyone.
Someone needs to pickup the original ideas of 1.x since they can't seem to stay focused, as their marketshare is ripe for grabbing.