The only way what you are seeing can be replicated is basically if the instrument is misused; i.e., it is purposely thrown out of focus or "de-focused". Would you call a family portrait picture that is grossly out of focus so everyone looks like blobs of color a good representation of the subject (or an instrument being used properly)? The fact that I can replicate your results by purposefully "misusing" an instrument actually is a warning that the bloated and boiling star images are NOT accurate representations of their true structure. You're basically just looking at the peculiarities of your own optics.
I disagree. It could be the best that camera can do. Over magnification is a real thing, I think it's caused by too shallow of a depth of field in combination with other things, but I'm not sure that is the case.
The P900 has manual focus, so first of all Tradman needs to try adjusting that. It could be that the focus adjustment resolution is not fine enough or is not capable of near infinite focusing distances. In my link on huge super zoom lenses, one of them could only focus as far as 30ish miles. I'm not sure why, but maybe the p900 can only focus as far as the moon or sun at full zoom.
As far as zoom goes, I have often thought that you could plot the focus adjustment of a telescope on a graph (knob turn distance vs optical focus distance) and be able to determine the curve, and use that to determine really how far the moon and sun are based on where the focus knob is adjusted to. Of course either a sun filter will be needed, or an indirect way of viewing it such as projecting the image onto a sheet of paper shaded in a box.