Gotta agree with Star18 on this one - cherry-picking is more "look at these 3 random games where they sucked (and ignore the other 79)!!" Star is applying a reasonable standard (teams over .600) and doing it consistently. It might be more reasonable to look at every game against +.600 competition, rather than just the last 14, or to compare their win % against top competition to those of recent champions to see if it's different, but it's definitely not an arbitrary selection standard.
It does look like a chink in their armor to examine performance against elite teams - unlike other recent teams with great W-L records, the Cavs basically never lost to a bad team this year (3 losses to lottery teams, 2 to Washington and 1 to Indiana who beat everybody), which helps to gloss over a more mediocre record against the relatively few elite teams.
Last year we lost against Charlotte, Washington twice, Toronto, Golden State, etc - all of these were mediocre teams at best. Against top teams we did very well, especially against the West. Maybe not coincidentally we kind of matched that in the playoffs, struggling against two teams with mediocre W-Ls and finding our stride against the two elite teams at the end. Cleveland has so far matched their pattern too, crushing two lesser teams but now struggling in their one game against a good squad. Whether that'll hold up remains to be seen though.
The Cavs have to be the favorites for the title now, but their struggles against top teams are very relevant to what's ahead of them.