I disagree that it would take anything like that long - with the advanced software tools and graphics libraries etc. that are now available, there'd be far less work on the fiddly graphics and UI stuff than in the past, the major work would be designing a good architecture. A bit like PCB layout really - with good placement, the layout's easy. Whether it would be financially viable is a whole 'nother question of course.
I suspect the devil is in the detail.
Just look at the other lower end basic packages like DIPtrace and Eagle, they are examples of maybe a decade(?) of development to get to their current point.
Altium itself took a similar length of time to get what most would now consider "basic" EDA functionality.
AutoTrax EDA (not the Protel one, the other one) is a benchmark example of what one guy can do on his own over maybe a decade or so.
Sure, it might be easier starting from scratch today, but I suspect it's impossible to get anything decently usable from scratch in less than a couple of years with a small team of programmers, say 3-4.
Sure, you could have something to show in 6 months, but it wouldn't be usable to anyone.
Dave.