As a teenager of the 80s and a potential listener to an all-80s format (never had it really tried here), what I heard wrong with all-80s formats was two-fold:
#1 - The format was always, ALWAYS slanted. That's great if you hit the "angle" I'm listening for; but how do you know if I want MTV-based playlists or the classic love songs of the 80s, or hairbands, or new-wave, or early 80s / urban cowboy crossovers, or... if you're going to be "80s" and appeal to a generic "80s" audience, you're going to have to go mass-appeal... and that means playing a bunch of records that don't test well, but people sing along with when they play. How do you test songs people deny they like but obviously DO when they are within earshot? I've never been able to figure that out; if you do, you'll likely make a mint!
#2 - The rotations are closer to current top-40 rotations. The "Oh, WOW!" becomes "oh... wow..." after hearing the same "lost oldie" day after day after day. (P.S. all-90s formats: you have the same problem!!!) You can pull the top 100 songs for each year 1980 - 1989 and have ONE THOUSAND songs that were big enough to not just chart but make the best of the entire year lists. Too big for you? Only take the top 50 and you're down to a (still large by today's standards) 500 song playlist. That still runs the risk of the same songs every day depending on your "hot" category size, but it's workable. Top 40 from each year = 400 songs... you get the idea. Please, though... if you're going to appeal to an oldies audience, DON'T only play 250 songs unless you're going to play them all in a row before you repeat each time, or maybe only have two categories.
#2.5 - If you're going to call yourself "80s," DON'T play 70s and 90s songs, please! If I tune in to the "80s" station (and I don't care that you call yourself "80s and more," you're "80s" to me and my target demo friends) and I hear something from '76 or '94, I'm gone and not coming back... EVEN IF I LIKE THE SONG. Why? Because you didn't deliver what you promised, and now I don't know what to expect from you... so I never crave you and never come back.
TIP: Take a hint from the syndicated "Retro Pop Reunion" and call yourselves "The biggest hits of the video music era"... OK, maybe choose a variation so you don't get a trademark infringement notice... but you get this idea.
I still think an 80s format can be successful, just like I think a 90s format can be successful (although trickier). I'm not convinced these formats will "plug-n-play" with traditional format layouts... although certainly experienced programmers can apply tricks of the trade to crafting a new production.
...It's ALSO not helped that many all-80s formats have been on rim-shot signals with minimal or no promotional $$$.
Two things make me think 80s radio can be a winner: #1, the tons and tons and TONS of online 80s stations being listened to by tons of listeners (especially at work). #2, so many other formats are filling their "gold" categories with 80s. Several stations locally which have lived in the 70s for a couple of decades have dropped much of that decade's hits and are replacing them with head-turning 80s hits. 80s music is used to sell all kinds of products on TV, and more than 50% of the retail establishments I enter are running an 80s-based or all-80s mix. If advertisers and retailers smell money in the 80s... why don't radio programmers?
One last thought: the 80s format is almost over; by that, I mean, the 80s target audience is getting ready to start moving OUT of the money demos. If 80s is going to be a big success, it's going to have to happen now, or it will end up being the 55+ audience relegated to failing AMs in not too many years.
The 90s format, however...
