The other issue that I need to consider is grounding the antenna...
There are multiple types of "grounds" that serve different purposes. In some
cases a "ground" optimized for one purpose won't work for a different one, so
you need to define what sort of ground you think you need.
The most common ones are:
1) RF Safety ground. This prevents a shock if the AC power line inside the
radio is shorted to the case, and some other conditions. Generally the third
wire on the AC outlet is sufficient for this. (It would have prevented my
making the patio door frame live with 120VAC to ground one time. Fortunately
nobody was hurt.)
2) Lightning ground. This diverts as much energy as possible to ground outside
the house in case lightning hits your antenna. Proper lightning protection is not
for the timid, but there are hams whose antenna systems are hit several times
a year with no damage to their indoor equipment. Whether or not you need
to worry about this depends on your antenna placement and the likelihood of
lightning strikes in your area.
3) RF ground. There are a couple variants of this. Basically it provides a
connection for the other side of your feedline: just as with a battery
circuit, you don't get any power to flow if only one wire is connected.
However, in the case of RF on coax, without an explicit connection to the
coax shield the "other side" of the RF will flow back up the outside of the
feedline. That can cause various quirks such as "RF in the shack", poor
audio, keyers with a mind of their own, etc.
A good RF ground is an important part of an end-fed wire antenna, though
the impact will vary depending on the relative wire length in wavelengths.
When the wire is 1/4 wavelength (for example, about 64' on 80m) then
maximum current is at the feedpoint, and any effective resistance can
greatly reduce the radiated power. When a wire is a multiple of 1/2 wave
(the same length of wire used on 40m or 20m) then the feedpoint is a high
impedance, where current is minimum. In this case, losses due to a poor
ground don't affect the efficiency as much. (This is why end-fed half wave
antennas are popular, at least for those who are NOT using a relay-switched
autotuner.) But at the same time the voltage across the antenna feedpoint
is higher. Still, if you don't provide a low impedance path for the "other side"
of the RF to flow, it will go back along the outside of the coax.
Making a low impedance "ground" connection isn't as easy as running a wire
to a ground rod, however, when the length of the connecting wire is a
significant portion of a wavelength. A perfect ground connected to the
antenna feedpoint using 1/4 wave of wire (about 8' on 10m) presents ahigh
impedance, and acts like no ground at all. Ideally for HF operation
the connecting wire would be less than 3' long, which isn't always practical.
The common solution is to connect a quarter wave wire to the ground side of
the coax at the feedpoint and string it around the room. You probably will need
multiple wires to handle several bands, though you might find that some aren't
needed (the 40m wire may work on 15m.) These are often called "counterpoises",
though that term is not well defined and gets applied to an assortment of
configurations. This provides a (relatively) low impedance so that most of the
RF flows into the wires rather than back along the coax.
At least that is a simple explanation. Whenever the topic of grouding comes
up you get a lot of disagreement, often because some posters are assuming one
or another type of ground.