Back in the 1990s, QST
ran an article about a pair of full-wave loops that hung from a common support, but were physically 90 degrees apart. The idea was to use a DPDT relay at the base to select one or the other. An option was to use a second relay to open the loops at the top for use on other bands. I can't remember the name of the thing so I can't point to the article, but that's all the info you need to build one. They were squares, but standing on a corner.
I built one for 30 M, and used a couple of nice ceramic-based relays for switching. Fed with 450-ohm twin lead. Worked really well on 30-20-17-15 until I got an SB-220... Couldn't use the amp on 30 M (200-W limit here in the U.S.), but I drove the antenna on the other bands with the amp, using a KW Johnson Matchbox tuner. One day I noticed the SWR seemed flaky and closing the top relay didn't affect tuning on any band. Hmmmm. Yep, the open-circuit voltage on at least one band was high enough to destroy a very hefty relay. It was a gift from my late Elmer, and I was pretty bummed out!
The moral is, "one size fits all" antennas can have unpleasant anomalies. A half-wave dipole has pretty high voltage at the ends; I've observed that with a 40-M inverted vee here, driven with a kW. I zorched some shrubbery when I placed one end in the bushes to conceal the rope I was using to tie it down to the anchor (temporary installation). The Burning Bush would have taken on a whole new meaning if conditions had been drier.
Moxon, a Brit, had fun building antennas with the ends folded back. If you do an Internet search for "moxon antenna" you'll see lots of examples. I think an antenna like this would be easy to model, too, and that would give you some insight into the claims of low radiation angle, etc.
Okay, the ol' brain hasn't completely failed. The QST
article was "Build the Versa Loop," and it was in the August 1989 issue (okay, I got the wrong decade....), p 22. There was a Feedback in the May 1990 issue concerning gain figures given in the article.