18) Every skill is relevant, as it adds to a player's ability to perform, but not all skills are equally important. It's never a bad thing to have high secondary skills, as they add no salary. You can ask yourself wether it's worthwhile to pay a premium for it on the market, or spend extra time training those skills though. It also depends on your team strategy; if you're employing an outside minded attack, Jump range on your center can add to your team's performance.
Generally you'll want high secondary skills as they make your player more efficient, but the question arises what to prioritize when you have limited resources (which is always). The resource training time usually dictates the amount of time you can spend on training secondaries; if you have a center with atrocious jump range, it will take very long before he'll be able to contribute to your team's outside offense. If he starts out with mediocre+, it could be wortwhile. I think you can discuss the ideal build for a player for a long time and not reach a definite answer.
Of course you want all skills to be as high as possible for your draftee, with maybe an exception to rebounding on guards as they do raise salary, and is the least effective on that position (still it adds to performance). Height does certainly matter, as it dictates what skills train fast and which don't.
19) Opinions differ on the matter of balance, and with the recent changes to the game this is surely to differ depending on who you ask. I'd say balance is important, especially in higher leagues, because opponents will exploit your weaknesses. If you have no outside offense at all, teams will use the tactics prediction (seen in scrimmages) and adjust their defense to gain advantage over you. Still it could be wortwhile though, as having no jump range on your guards lower their salary, so you can have higher defense for the same salary, compensating for their gained advantage. For a long time, look inside strategies have been dominant, so there will be plenty of people who will argue balance is not needed. This is because the game engine (AI) doesn't really adapt well to it, because if you use a 2-3 zone, the passing lanes will open up so you won't get much advantage. However, with the new features of being able to predict your opponents strategy, this might change.
On defense, it's the same dilemma you face. You can indeed choose to pair another offense big, giving you an offensive advantage, but leaving your defense weak. It's not very easy to say what's better, and that's the beauty of a strategy game; there are multiple valid options.
20) Indeed it doesn't cost you anything to place players on the transfer list, and certainly there's money to be made working the transfer market. However, there are some downside on day-trading; players drop in game shape when they are purchased, so they'll be less effective to start out. Also, there is a transfer fee that is highest when you just bought a player, and will gradually drop to 3% after you own the player for about 1 season.