The murders not being solved aren't "Perfect crimes". It's a matter of resources and timing.
99% of murder investigations mostly boil down to:
1. Literally witnessed by one or more people, possibly officers or on camera.
2. They brag about it.
3. A brief investigation of friends and family where it turns out so and so always hated them and happens to have a gun, and hey look at that the ballistics match.
There are a very small number of officers in comparison to the total population (as it should be), and the vast majority of them are not the kind being assigned to homicide.
Some major % of the "unsolved" murders in the US that mostly just get thrown on the pile because they find out 2 weeks later and either can't identify the victim, or can't find enough useful information to start investigating. Forensics is very useful, but hardly as portrayed in shows like CSI, and the simple realities of "well we didn't find much, found out weeks later, and it'll be weeks before we get any lab results back" often just mean there's 10 other "no shit" murders to deal with instead.
And this isn't even on the core topic here of "could this have been a hit by Boeing", which is just insanely unbelievable, ESPECIALLY given the method. People are off handledly mentioning things like Ricin or Polonoium attacks, but the important things about those is that they are INSANELY LETHAL and extremely easy to control.
"Lets infect him with pneumonia that turns into MRSA" has got to be one of the most risky, difficult, expensive, and unreliable methods of killing someone ever.
Hell if you want deniability there's probably at least 10 or so ways to easily cause a human to have a heart attack and look like they died of natural causes as compared to some magic MRSA gun.