As my answer is quoted in the OP's question, let me elaborate.
Specifically on my T-antenna with auto tuner at feed point/base, as the OP left a recent comment asking about this antenna, hence the focus of this antenna in this answer.
It is a T-antenna, which means a vertical radiator 10m high, and a top-spanning 30m horizontal at each end, originally built for 160m and 80m. It resembles a T, hence I call it my T-antenna, I do believe this is an accepted term though.
Using it over the last 3 years, I get reasonable DX performance for 160m and 80m. Due to the vertical aspect of the antenna.
However the auto tuner is capable of matching this antenna for 40m, and I discovered that it has some NVIS properties for 40m, this is probably due to the 2x 30m horizontal legs of the T-top.
Discovered by accident, but it did allow me to use this antenna for NVIS 40m to the extend that I do not use my horizontal dipole all that much anymore.
However, it still spans 60m horizontally, which is actually more than a 40m dipole. So if you are space restricted, this would not be an ideal antenna for NVIS purposes. Its main purpose is still 160/80m DX.
Just to be clear, I built this antenna as an experiment, and decided to keep it. There were no scientific designs involved, nor modelling. I could go up 10m, so I did, I could go wide 30m at each side, so I did. I put an auto tuner at feed point, and I played radio. That is all, so if looking for a scientific answer for T-antenna's which are designed for band X, but have properties on band Y, then this is not your answer. For more scientific calculations of a T-antenna, you can see this page although originally written for the 136kHz band.