So what is an ideal Lp(a) level to reverse arterial plaque? The body of research on the subject is not as big we might like, but let’s look at the key studies to date:
1. Lp(a) < 10 mg/dl. Medscape states that “many US lipidologists generally consider an Lp(a) level of less than 10 mg/dL to indicate a lower cardiovascular risk. levels higher than 10 mg/dL are associated with an increase in cardiovascular risk (see Interpretation).” 
2. Lp(a) < 30 mg/dl. One fairly large and recent study from Europe noted that heart attack risk was not significantly affected until levels of Lp(a) exceeded 30.  In their opinion doctors only need to be concerned with extreme levels and not “normal” Lp(a) levels. You can read the whole study in pdf form here.
30, by the way, is the top of the range for most labs from what I have read. So it may seem a bit lazy for anyone to recommend that as a target value! However, a point that is often missed is that values vary GREATLY across the population. A siginficant percentage of the population (15%) has Lp(a) between 30 and 250!  Therefore, to many people out there, 30 looks like a darn good number!
4. It Depends on Your LDL-C (actally LDL-P)?? There is a theory out there in the alternative medical community that Lp(a) is THE risk factor for heart disease and arterial plaque. According to this way of thinking, LDL, small and large, is only a bit player : Lp(a) is the real culprit. And they have some very convincing arguments.
In addition, there was a study that actually found that for lower levels of LDL-C (<70), there was no additional risk added from Lp(a). Here is the study that made that point:
“In contrast to these variables the association between Lp(a) and coronary artery disease was dependent upon the serum concentrations of LDL-cholesterol, -Lp-cholesterol and total cholesterol. At concentrations below the respective median for each variable, odds ratios of between 1.42 and 1.67 were calculated whereas at concentrations above the respective medians the odds ratios ranged from 4.50 to 6.33 (P < 0.001). Our data, therefore, suggest that increasing LDL concentrations markedly increase the risk of coronary artery disease due to elevated Lp(a) levels.” 
NOTE: This is a less than desireable approach in my opinion, because they really should have used LDL-P instead of LDL-C.
Now I do not think one should think this way. Why take a chance? In my opinion, if you have high Lp(a), you should try to lower it as much as possible. Dr. Davis argued strongly that high values of Lp(a) will add plaque to just about anyone and is responsible for most heart attacks in younger men. So I think the consensus now is that you have control both LDL-P and Lp(a).
CONCLUSION. One should try to get Lp(a) below 30 mg/dl and realize that 10-30 is a grey area that needs more research. The men most at greatest risk are those very high Lp(a) and it would be extremely difficult for them to hit the 10-30 mg/dl range anyway. There is also some evidence that LDL-P and Lp(a) work together to increase arterial plaque an also that Lp(a) is a completely independent risk factor for heart disease.
2) Circulation, 2008 Jan 15, 117(2):176-84, “Extreme lipoprotein(a) levels and risk of myocardial infarction in the general population: the Copenhagen City Heart Study”
4) Circulation, 2001; 104:1108-1113, “Coronary Heart Disease Prediction From Lipoprotein Cholesterol Levels, Triglycerides, Lipoprotein(a), Apolipoproteins A-I and B, and HDL Density Subfractions The Atherosclerosis Risk in Communities (ARIC) Study”
5) Atherosclerosis, Dec 1986, 62(3): \249-257, “The association between serum Lp(a) concentrations and angiographically assessed coronary atherosclerosis: Dependence on serum LDL levels”