My first suggestion would be to use a terminated folded dipole cut large
enough for 80m and the rest would just work. Have a look at Mr Cebik's
website for details.

http://www.cebik.com/wire/t2fd.html
http://www.cebik.com/wire/wbfd.html

Your idea for the design is the kind of setup that would be easy to
model in Eznec/4nec2 etc. I did this once (only) for a 10m/6m design 90
degree offset inverted V. I noted that there was a much narrower SWR
bandwidth on the 10M side. I didnt delve into why because I never built it.

Apologies for not answering your question directly though.

Cheers Bob VK2YQA

C. J. Clegg wrote:
-----

Questions...

Is the standard 468/F formula likely to be anywhere near accurate, or
will the three antennas interact with one another to throw the
calculations off?

In particular, is the 4.5-MHz frequency too near the 80-meter band
such that interaction will be greater than otherwise?

If the interactions are sufficient to throw the formula off, will my
antennas end up being longer or shorter than the formula length?

Thanks.... CJ