This is a woefully incomplete analysis. This is likely largely the fault of poor reporting of homicide details or at least no uniform data structure that makes it easy to aggregate categories between cities, but most homicide is domestic. The obvious hypothesis as to why homicides increased while burglary decreased is people were home more and under a lot of stress. Thus, they're less likely to be burglarized because their houses aren't empty all day, but they're also more likely to murder each other. Proactive policing can't stop this unless the police are randomly going into people's houses.

This isn't necessarily the actual explanation, but you can't decide between this and competing hypotheses without analyzing the nature of the homicides that happened. Were the increases mostly due to gangs and ruffians shooting each other on the street? Or were they mostly due to people at home shooting their spouses and kids?

Don't forget the huge jump in unemployment as well. Unemployed people tend to commit more crimes. So why not burglary? Well again, if people are home, it's a lot harder to burglarize them. Why not drug crime? A whole lot of drug use was legalized in the past year, so you don't want to look at drug crime, you want to look at drug usage.

There's also hospital overcrowding. An increase in murder can just mean the same number of people shooting each other as normal, but more of them died. Did attempted murder also increase? I don't even see that question asked.

I obviously don't have any answers because I haven't done the analysis either, and I'm not going to because I'm not a journalist or analyst and already have a job, but the people whose job is to do this kind of analysis should at least do it correctly.