In my lab, we have never experienced any effect of cell density of H9c2 with hypoxia. 

Logically, the higher is the cell density, the higher is the oxygen consumption rate per ml of medium. Everything else being equal (type of well, thickness of medium), your cells may therefore be more likely to trigger HIF expression at 2 to 5% O2 (incubator atmosphere) if cell density is higher, because the pO2 at the bottom of the well will be lower. We actually observed this a few years ago using clonal INS1E beta-cells. In contrast, I don't think it should make a difference if you use 1% O2 or less in the incubator.

I have no experience with your cell type.

According to my experience in hypoxia more cell density associated with high expression of HIF in 5% CO2 and 1% O2 atmosphere.

I have no experience with your cell type.

I have no experience with your cell line but cells which are overly confluent come out of log growth phase and often gene expression changes. We have found differences in plating efficiency in hypoxia experiments if the cells are too confluent ( in our hands hypoxia has less effect on cell survival) For experiments i would always keep your cells subconfluent at around 80% confluence max. Similarly do not use your cells at too low a plating efficiency as this too stresses the cells and can change results.