Your findings support exactly what I'm suggesting could be a quality issue - the fact that the sweet spot for quality falls at the highest bit rate but only a medium frame rate suggests that the recorded picture is bit rate starved at the higher frame rates with great compression being applied and thus more artefacing and loss of quality. Otherwise were it sufficient, the picture at 30fps would be the same quality (individual frames) as at 15fps in the day with the effect that overall quality video would remain the same (but smoother at 30fps).
If the bit rate were sufficient, it should be possible to get a top quality picture at the full upper frame limit of the camera (at least in daylight - obviously at night other factors come into play which favour lower frame rates such as light gathering ability).
Yes, it seems to be a race to the bottom.
if you are a business and want to save months of video streams from 16 or 32 cameras, or are on a budget and don't want to install (what is really enough) hard drive capacity, then the are advantages to having the smallest possible file size.
However, compression is never a good thing and I personally would like to see the user given the option on the compression front so the user can choose what is more important to them, max quality or minimum file size.
That's the impression I get which is actually a real shame for the 5231. A 25x zoom should be compromised on quality from the start, because of the number of groups of lens in use - the old photographers rule here of you usually get the best quality picture from a fixed focal length lens comes into play, and not from a zoom and especially not from a super zoom. However, it appears the glass quality of the 49225 my be far superior to that used in the much smaller focal length 5231. This is a pity and a real limiting factor given that sales of the 5231 are likely to be much higher due to smaller size, a lack of need for a motorised turret in many applications and cost.
I don't mean to appear rude Fenderman, especially as I'm new here but you seem to not properly understand how compression works. As I said above compression works by referencing the change between frames in the picture. This means that the amount of compression applied is proportional to the amount of change in the picture from one frame to the next. It's also proportionate to the amount of detail. Thus any movement in the picture (be it object or camera), the greater the positional change of pixels within the frame and the higher the amount of compression needed to record the change - the bit rate fixes the available space so all the detail of the picture has to be fitted within the fixed bit rate. The more data there is, the less it fits and so the higher the compression applied. The more compression, the more pixels within the frame become referenced by areas of similar colour / texture rather than by their true content, and the more errors, artefacts and loss of definition within the picture (including readability of text / blurring of moving areas etc).
It's also a common misconception to believe that to double the quality you have to double the bit rate because with the codec only referencing the change from frame to frame, it's only part of the full frame that's being compressed so it doesn't require double the space to 1/2 the compression rate.
There's a good simple article here that specifically explains how a higher bit rate is needed for moving objects - the example here is a static CCTV picture vs one at a busy intersection (where obviously there's a lot of traffic and thus movement within the picture):
CBR vs VBR vs MBR - Surveillance Streaming
