Fair enough.

1) I would say that that weird, counterintuitive, and non-obvious goals are those that would be most likely to benefit from lying. This is independent from whether these goals are worthwhile.

1A) I would agree with the OP that it is more likely for worthwhile goals to benefit less from lying, as it should ideally be easier to convince people of their benefit truthfully. However, considering the amount of irrationality in the world, there will be plenty of goals that are worthwhile and would not be easy to convince people of using only truthful methods.

2) Lying may be a more efficient way to accomplish worthwhile goals than not. By denying the tool of lying, one's options are limited.

2A) Now, this does not mean that lying is a worthwhile tool to use. It may be more efficient, but also more risky. Likewise, some goals may be undermined by using the method of lying to get to them. A rational agent is best off calculating the expected utility of lying as a tool before using it.