Logitech, Asus and Razer aren't going to sell very many boards no matter how great they are at $400+, part of the appeal of customs is that it's unique or at least low volume which means even fewer sales for these companies. Worse still it's a fickle industry that shifts with the wind and they can't pivot fast enough.
At that level, 2 years to develop a new product is not out of line (and can go up, rarely does it go down), between design, approval of execs, prototypes, testing, changes, durability testing, another round of approvals and then final manufacturing, which could take a while as they would want large numbers and you simply can't CNC a case even remotely as fast as injection molding. A nice CNC case can take hours, an injection molder can do thousands or even tens of thousands in that time.
A group buy ignores a lot of this process, it doesn't have a chain of command making time consuming decisions, and a bunch of bean counters working on that, not to mention the design is usually at least somewhat fleshed out before it's even presented (which could take weeks or even months of CAD work to get how they want). They also don't need the numbers because they aren't in it for the profit and they aren't dependent on that profit, no one is being paid overtime while the group makes changes or decisions and no money coming in.
In a group buy if you want to add an extra screw for stiffness, add it, it's 100 screws and a few minutes in CAD, that's pennies for 100 keyboards, if Logitech wants to add a screw to an order of 50,000 keyboards now you need to justify it to bean counters because you just added another round of design, prototypes, testing, and committed to buying 50,000 more screws. That one screw could add several dollars to every keyboard and that's several dollars not going to shareholders. Don't forget, Logitech has distributors and outlets who cut the profit in half so any time the price goes up, it goes up twice what it should. A keyboard you paid $300 for on a group buy would need to probably sell it for $500 and be engineered down to $300 just to make it make sense. You can engineer things to be cheaper but cutting out 50% is difficult even with quantity.
Ultimately though it's just numbers.
Would you rather sell 5000 keyboards for $400 and make $20 per keyboard or would you rather sell 500k keyboards at $40 and make $10 each?. There's almost no profit in that 5000 boards by the time you deal with warranties however there's plenty of profit and room for returns on the 500k boards. This is why their keyboards are always conservative and never really cutting edge in terms of features and form factor, it limits sales, and worse, what happens if you get it wrong? It's a keyboard, how wrong can you get it? A group buy has pre-purchases lined up, imagine Logitech coming up with a new form factor, say 73% and spends tons promoting it and it flops. A 100% board will never flop, you can always just price it down and sell it to offices or just let it sit on shelves until it finally sells out. That 73% could sit for decades trying to move all of it.
This is probably why they developed their own switches.
Form factor can be copied, everyone has MX switches, so how do they differentiate themselves enough without massive risk? With a switch no one else has. This is them being "innovative" without taking a ton of risk. Sure the switch was costly to design and manufacture but it was developed with Omron, if it fails, Logitech can just push those keyboards down their lineup and blow them out to normies. Omron could also start to offer the switch to other companies to recover their costs. It's high investment for Logitech and Omron but has relatively low risk compared to trying to design something truly innovative and compete in a fair yet niche market place with other MX switches where they simply couldn't compete and make the money they require.
Don't forget all of these above also ignores color variations (all of which would need durability testing and approvals and surveys to decide what will sell best), the need for 10 different switches, and of course boxes to match, all of which ads multiple product SKUS and tons of logistics department. This also means stores would need to decide which to carry and which not, something Logitech has little control over. Logitech could find that people for some reason really loved the green model with MX clear but stores only wanted to stock black with MX blue to sell to mainstream users leaving Logitech to deal with all the green orders themselves and not enough inventory of black with blue switches.
Basically way too much risk, way to little profit and far more complicated than they would want to deal with when they can keep doing what they do and rake in the money.
