There are a couple of things going on here. One is incumbency advantage. The Republican wave in 2010 captured a lot of governorships, and to some extent the incumbency advantage has carried them through. For states with a two-term limit, governors elected in 2010 are just now reaching their term limits; therefore, this will be the first election since 2010 in which they haven't had the benefits of incumbency.
The second factor that abets this is the fact that partisan lean is much weaker in governor races than it is in in races for legislative offices. The 2016 the Cook Political Report had this to say on the subject:
The correlation between presidential performance and which party wins a statewide race isn’t as prevalent in gubernatorial races as it is with U.S. Senate contests
A discussion in a recent FiveThirtyEight podcast put the effect of partisan lean on governor's races at about 1/3 of what it is in legislative races. In a recent piece on split-ticket voting they observed:
Incumbents tend to win re-election, even if their political party does not match the one their constituents generally prefer.
You can see this difference playing out in states like Maryland, where incumbent Republican governor Larry Hogan appears to be poised to cruise to an relatively easy victory, while Democratic Senate candidate Ben Cardin looks practically guaranteed to win his race. (Note that these two races have exactly the same constituency, namely the entire state of Maryland.)
The reason for this difference is thought to be that governors are perceived (probably correctly) to be a lot more independent than legislators. Voters judge that candidate for governor from "the other" party who promises to govern as a moderate can possibly make good on that promise if elected. On the other hand, if you elect that same candidate to the US Senate, no matter how moderate they personally might be, they are going to be under a lot of pressure to fall into line with their party's leadership, which is probably anything but moderate, if they are elected.
So, these two factors explain much of the Republican gubernatorial dominance: more willingness by voters to cross party lines in gubernatorial elections than in legislative ones, and the knock-on effects of some very good years for Republicans in the early 2010s. This year looks like it will see some reversion to the mean. The 2018 FiveThirtyEight gubernatorial election forecast predicts something close to parity for Republicans and Democrats after the next election. Still, you will see some of these effects in force, as Republicans are expected to hold on to states like Massachusetts and Maryland, both of which are otherwise solidly Democratic.