The Megaprocessor uses through hole components. Something similar using surface mount components could be substantially cheaper, mostly because it could be smaller. This means it will need less of all the expensive supporting parts. The components themselves are only a small part of the cost.
"How big would a modern CPU be at this scale? The Apple A8X, found in the iPad Air 2, contains about 3 billion transistors. (This is comparable to the number of transistors in modern desktop computer CPUs as well.) At the scale of the MOnSter 6502, that would take about 885,000 square feet (82000 square meters) — an area about 940 ft (286 m) square."
Follows the same logic, but much cheaper.
A good programming analogy is people confusing the purpose of a programming Koan collection and helpfully suggesting the participant should google the answers. Or even worse than programming Koans, ditto but for meditation Koans. The point of the experience is to do it your way, by yourself...
Personally I'd like to play with RTL gate logic using modern transistors "in my infinite spare time". There's a certain alternative history component aspect. Imagine if instead of leaky, low beta, low Ft, low breakdown voltage germanium transistors, RTL was implemented with top of the line 2010s discretes. Yeah yeah I know RTL begat DTL begat TTL using contemporary technology, but in alternate history world, how would life have changed?
And you still get blinkenlights per gate or transistor in emulated design.