The term highlights has a longer history than one might think, given that its meaning today is so closely related to radio or television. Etymonline, glossing over the pluralized version (which it doesn't distinguish as a headword), says
highlight (n.) 1650s, originally of paintings, "the brightest part of a subject," from high (adj.) + light (n.). The figurative sense of "outstanding feature or characteristic" is from 1855.... Related: Highlights.
The Collins Online site avoids this dilemma (geddit? LEMMA), even giving it its own frequency graph:
In the words of a comment I made recently to the Collins Online site
The definition 'a selection on the TV or radio of the most important and exciting parts of an event, esp a sporting event' doesn't work any more. To judge by the BBC's coverage of Rio 2016, "highlights" seems to mean "about an hour of celebrity chat, punctuated by very occasional and sparse clips of sports action".
<rant id="2" ferocity="mild – not even a rant really, just an occasion of vague regret and nostalgia">
I know I know I KNOW, this is the way language develops – I've defended so-called "mistakes" often enough in this blog.
But I'll never say (or write, except here, of course) appeal the decision. The most recent "infraction"("He only does it to annoy, because he knows it teases") was probably to do with drug cheats before Rio. I compared this construction (and the version we dinosaurs still use, with a preposition and no direct object) in the British National Corpus and in its American analogue COCA.
The search appeal against the [n*] (by the magic of BNC, you can just click on that link) occurs 136 times in BNC. Meanwhile, appeal the [n*] occurs only 36 times: the version with the preposition outnumbers the newcomer about 4:1. (That word newcomer suggests a possible PhD study: "The usage of non-traditional grammatical forms – an age-related study". That would put some numbers on what to me at least is a self-evident truth: as language develops over time, the trail-blazers are the young.)
In COCA, unsurprisingly (although possibly the extent of the preponderance [nearly 20:1] is a bit of a surprise), the relative weights are reversed: the American English strong preference is for appeal the [n*]. (Given the state of the hedge  I must leave the workings as an exercise for the reader.)