Assuming a 300' loop, at low frequencies up to ~3-4MHz it will likely model as more of a blob shape with more "gain" (actually less loss) straight up than towards the horizon. It should be a decent NVIS antenna.
Around ~6-7MHz is going to shift the pattern more towards the horizon than straight up. Not exactly a big deal IMO, as NVIS tends to roll off above 7MHz, if even that most days, anyway.
As you go higher in frequencies the pattern towards starts breaking into multiple lobes, with the resulting gain and loses depending upon the distribution of those lobes.
Losses will increase as the loop approaches ground, especially as frequency drops. Works pretty much that way for most any horizontal antenna, be it a dipole, loop, etc. That said a low loop will likely have ground losses across pretty much the whole HF spectrum, thus the previous mention of "gain" in quotes.
Actually, lots of us have receiving loops placed directly on the ground.
Losses can be offset with a preamp if truly needed, though a preamp is often not a requirement at low frequencies, especially with large loop sizes. My 148' LoG typically suffices without a preamp from MW up to around mid-HF.
Summary? A large loop even at relatively low heights should suffice as a decent receiving antenna where losses at higher frequencies can be offset with a preamp if needed. Conversely, it not so much as a great transmitting antenna, though lots of people routinely work lots of contacts with horizontal antennas at low heights. YMMV.
Now as compared to an 80m dipole at 35'+ height? Ground losses still will be present at low frequencies, but that issue should improve around ~7MHz.
The dipole is still low enough to more favor NVIS at lower-HF frequencies. As frequency increases above ~7MHz the dipole should start having more gain at angles more towards the horizon than straight up.
As to pattern the dipole is going to become more bidirectional as frequencies increase, though opting for an inverted V deployment will somewhat improve an omnidirectional pattern at the sacrifice of some broadside directivity.
The dipole will start breaking into more lobes as frequency increases, too.
Personal thoughts? Given the feedpoint height differences present, I would likely favor the dipole with a feedpoint at 35'+..... though perhaps a different type of dipole with symmetrical balandce. An 90'+ doublet center fed at 35'+ as an inverted V should make for a decent HF antenna, especially from 80m to mid-HF. Coaxial feedline would suffice for receiving; just ignore the mismatches. If transmitting is desired, feeding a doublet with ladder line and using a tuner would be the (much!) better approach.
https://ftp.unpad.ac.id/orari/library/library-sw-hw/amateur-radio/ant/docs/Introducing%20the%20All-Band%20Doublet.htm
Topic: Sky loop...height vs diameter (Read 5920 times)
