I thought that was the Ford GT’s job
You can have two halo cars - but at least one of them has to bring people into the showrooms. The Ford GT isn't all that effective at that - it's more a corporate bragging vehicle, honestly. It would be more effective as a customer draw if they'd made enough to put one in every Ford dealer in America plus satisfy all the other demand. They didn't, so it's not the kind of halo car that will draw in the married person who dreams of getting one, but ends up driving out in an Edge or Explorer because they have three kids and have to get a practical car for that reason.
Nissan has a similar situation - the 'unobtainable' halo car is the GT-R, the one that's on every dealership floor is the 370 (soon to be 400) Z. The GT-R gets the customers interested in the brand, the Z brings them in to the showroom - and then they (statistically) drive out in a Rogue or something.
Similar reason is why Sir William insisted on there being the ungainly 2+2 E-Types - guy would come in to drool at the 2 seater E, but couldn't get the wife to agree to anything less than a four seater. He'd drive out with a compromise, the 2+2 - an E-Type for him, but four seats with room for the shopping and kids to make the wife happy. Some of the circumstances have changed with society in the intervening half century, but the basic mechanism still works.
The point still stands on manual vs convertible vette. Take rate for both is low and IMO the latter isn’t even necessary considering all “hardtop” models are actually targa
Take a look at the price differential of automatic vs manual vice hardtop vs convertible. People are not willing to pay a lot of money to get a manual version (in fact supposedly they complain if it's not zero cost or just a token amount) but they will pay stupid amounts of money to get a folding roof. $30M out to just make a manual option that will get you $10M back at best doesn't really make a lot of sense. Consider that the Mustang sold more copies in just the first two months of this year than the entire MY2018 production run of the Vette and you might begin to see the difference that volume makes. Assuming a similar 20% take rate and a nominal $1000 upcharge for the manual Mustang (not based in reality, just making math easy) the Mustang will have completely paid off its $30M manual trans development in just two years.
Also, keep in mind that the Mustang is a conventional RWD car - the $30M is assuming you can just call up a bunch of vendors and have them send over their most appropriate off the shelf offerings; you can still do that for conventional front engine, mid trans RWD. For the C8, they said they couldn't get any off the shelf manual mid engine gearboxes from suppliers and would have to have designed a new one from scratch, greatly increasing the cost - which could not have been shared by anything else - and consequently making the break-even for a gearbox even more difficult.
GM wasn't kidding about not being able to get MR manual gearbox/transaxle setups for longitudinal engines. This has long been a problem for MR kit car builders (and even outfits like Lotus and Noble); the last I looked (which admittedly was at least five or so years ago), most of the kit car solutions were things like "adapt a 928 transaxle" or "adapt a Porsche 911 transaxle". There were a couple of custom makers but they wanted $25K or more per copy and clearly weren't going to be able to really mass produce their designs as there was far too much hand fitting required. And of course, Porsche isn't about to sell or license their transaxle designs to GM.
You may also want to look at how much it costs just to tool up to make an all new transmission design:
https://www.fool.com/investing/gene...nd-general-motors-teamed-up-on-transmiss.aspx