You definitely should be convinced that it causes heart disease because it doesn't. However, just because it's poorly understood doesn't mean that it's negligible. A correlation is still a correlation and is considered important if it is significant. An appropriate analogy would be that many drugs are not understood and the mechanism is relatively unknown. However, that doesn't mean they cannot be used if their safety has been assessed and the benefits outweigh the detriments.
Granted, I would like further research to be done, but there is sensible physiology behind it, especially if you get everything literally measured, as opposed to getting things like "calculated LDL."