Re the question of whether grounding attracts or rejects, I assume this refers to lightning. In MY opinion, there is no question that grounding a tower or (if it were possible) to ground a dish, that it would make a lightning strike LESS not more likely. This is because the likelihood of a lightning strike is usually dependent on two things, the difference in voltage between the cloud and tower, and the distance between the cloud and tower. These two things basically end up giving a volts per inch type gradient, that will end up determining whether the breakdown point of the air has been exceeded, ie the lightning strike won't occur until the air kind of breaks down and becomes conductive. Grounding will effectively cut the voltage differential in half, because when a highly positive _OR_ negative cloud passes over, the poorly conductive soil surface will become polarized to the opposite charge. Ie when the tower is grounded, the tower will be at zero potential vs say the +V potential of the cloud, whereas an ungrounded tower will effectively be -V vs the +V potential, so the voltage difference will be 2V instead of 1V. People talk about giving the lightning a path to ground, which is an issue once the strike has started, but relative to the odds of having a strike, grounding will make the strike less likely, because it keeps the tower at zero potential rather than developing a polarized charge opposite to that of the cloud.
However, that is only part of the problem. Other important issues are (1) protecting your house from burning down if there IS a strike, (2) protecting your electronic equipment not from a strike, but from the large static charges the build up prior to a strike, and (3) protecting your equipment from voltage differences that develop between the ground out at the dish vs the ground in your house, and (4) the question of if you ground, WHERE do you ground.
The NEC is mainly concerned with protecting your house, not your electronics, and is thus mainly concerned with dissipating a lightning strike once it has occurred. While logical, it may or may not protect your electronics, depending upon how it is interpreted.
The NEC is pretty clear that any antenna system such as a dish or TV antenna should be grounded at your service ground where the power comes into your house. Ie this is NOT what some people do, ie trying to actually ground the dish itself. While I think grounding the dish itself might make a strike less likely, unless it is done properly, which involves HUGE grounding straps that connect the dish ground to the house ground, which for most installations might cost thousands of $s, would make things worse than not grounding the dish at all. Basically you could end up with a different ground potential out at the dish compared to what you have in the house, and this can end up causing all sorts of issues when you have an lnb whose ground might be greatly different from the ground of the receiver. Basically, it's better to ground the shield of the coax where it enters your house, and this should be done via the house service ground.
The problems come up with situations where there is no convenient way to route the coax so that it enters the house near the service ground. People tend to bend the rules, and the question is which way of bending the rules is best. I don't recommend doing what I do, so I won't mention how I've bent the rules, but I will say that the ONLY equipment that I've had killed by static discharge have been receivers that were directly or indirectly connected to service grounds that were away from the house. Ie I went for years without a problem, until I let DirecTV do a free "upgrade" to give me local channels. They installed their dish out by the power pole that feeds my house, and they grounded their dish on the power pole ground, instead of at the service ground at the house. I think I lost 4 DTV receivers and one other receiver sitting next to the DTV receiver due to that ground. I've since disconnected their ground block, and haven't had a problem since. During the period that I lost the 4 DTV receivers, I had NO problems with any of my FTA dishes that were ungrounded.
Bottom line..... If you ground, ground to the house service ground, or at least to a solid ground connected to this. Do NOT ground AT your dish with a ground separate from the house ground.
Now, re spark gaps... I really think that these are more useful for open antennas like a TV antenna, where the actual driven element is exposed to the lightning or static charges. I've seen these mostly used for balanced feed lines, like TV twin lead. I also believe that once potentials are great enough for a spark gap to function, you've got enough potential to fry most electronics anyway, so it's mainly a way of protecting your house.
Since the driven element of your sat system is enclosed within the grounded case of your LNB, I don't see any convenent way to use a spark gap here, unless it is between the shield and some ground.
Just my opinion, which I've given many times, and each time I've received lots of opposing opinions, and I don't expect many to agree,...... but you asked for opinions.
