We can speculate all we want, but the actual decision belongs to the people who choose to design and fabricate the next generations of space hardware.
My bet is that they will choose to do it for the lowest cost that is compatible with acceptable probability for mission success. And, the cost of programs that are unnecessarily complex comes from the extra personnel costs for the long design, fabrication and testing processes required for the additional complexity. That was certainly true for the Webb telescope, which was long delayed. And, part of that delay was waiting for a rocket with a large enough payload fairing and enough reliability - which eventually turned out to be the Arian launch vehicle. Still, while on its way to its mission location, there was much nail-biting as it automatically assembled itself into mission-ready configuration.
The LEO assembly option bypasses much of the complexity for getting it into mission-ready configuration, because it can be launched in multiple segments that each can fit into available launch vehicle fairings and be easily assembled once of the atmosphere. And, it can be tested in its final mission configuration before being sent to its mission location - with repair if needed while still accessible in LEO.
BTW, "billionaires" are not the ones who are going to be doing that work.
But, there are some who intend to go into space. And Jared Isaacman has already done it, including going out into the Van Allen Radiation Belt and doing EVA. See
https://en.wikipedia.org/wiki/Jared_Isaacman .