I have been thinking how you could recreate the "Great Kit" experience today, and it is very hard. It is not the same as building a kit from a magazine project, it was something completely different.
You were led part by part through the construction, so novices didn't have to panic. All along the way, techniques were clearly explained. A full stepped procedure for powering up, testing, debugging and calibrating was included. Finally, there has to be some kind of service available for the people who stuff up so badly that their kit will never work. As well as being a kit, they were a mini university course as well.
Heathkit and others like Sabtronics, Sinclair, University Graham in Australia and many others were able to make these "Great Kits" because the prices of the equivalent products was so high. Production cost markups were just so big, that it was cheaper to sell a kit. People built the University Graham VTVM (Vacuum Tube Voltmeter) in the 60's because it was much cheaper then any built meter of the same quality. Until FETs became available for meters, you had to have a VTVM to get the 10M input impedance you needed.
Sinclair in its Scientific Calculator kit around about '74 - the first affordable scientific calculator - were able to offer a 2.50 Pounds repair option to anywhere in the world - so obviously when you paid for the kit, you were really paying for a second calculator board as well that Sinclair would send if your attempt failed.
The only successful kits I can think of today are the very specialized custom projects like the 3D printers, quadrocopters, and home DNA sequencing kits where there are no cheap commercial products available right now.
If anyone can think of a business model and a product range that would work today, I would love to know.
Richard.