For quite some time I have had a theory about the causes of uneven or chalky results when retrobriting plastic items. These basically boil down to two causes.
1: Uneven application of the Retrobrite mixture, especially if applied in the form of a paste and even more so if cling wrap is applied to prevent drying.
2: The plastic surface is already damaged. By this I mean that the exposure to heat/light/oxygen that discoloured the plastic in the first place also affects the physical and chemical properties of the plastic, causing the uneven results.
For a long time, I had prevented the first problem from happening by completely submerging the item in a tray containing a laundry whitener solution containing sodium percarbonate. An example of which is shown on the left.


In the picture on the right, I'm soaking the top of a VIC-20 case that had been heavily yellowed. I used a shallow tray large enough to hold the entire item. Enough water to completely submerge the item was added, and a cup of the laundry whitener was added and stirred in. The item may need to be weighed down to ensure full immersion. The tray was then left in the sun for a few hours, stirring occasionally. With the harsh Australian sun, I was able to get full results within 3 hours. On a cloudy day, it took longer. I am of the opinion the milky solution actually helps disperse the UV light, helping to give an even result. The suds on the surface had no effect on how even the results were. In any case they dispersed after a while.
I originally whitened this VIC-20 case back in 2011, before and after shown below. It should be noted that the case was originally orange all over, I had already done some testing on the right hand side, which is why the yellowing looks uneven.
There was some chalkiness on the finished case, except where there had been a label, which I had removed before I started soaking it. The plastic was not chalky where the label was, and this prompted me to think about the damaged plastic scenario, which is the subject of this post. As I wasn't using this VIC-20 case for anything, I stored it in a dark place.
After several years in storage the case had yellowed again, though nowhere as badly as it was originally. Of course while Retrobrite reverses the yellowing process, it does nothing to prevent it happening again. I decided to repeat the process, but this time I wanted to see if Retrobrite actually causes damage (chalkiness, brittleness etc) as some people think. So this time I decided to soak the item for two weeks in the harsh Australian sun. On the left is the VIC-20 case showing the yellowing that has returned after the original treatment in 2011. On the right is the same case after the two week soaking.
The rationale for my experiment is to see if my theory that existing damage to the case causes brittleness, unevenness or chalkiness. To provide a control, I had removed the label, leaving an area of plastic that had never been exposed to oxygen or light, and had never gone yellow. The inside of the case provided another control, only this time it has been exposed to oxygen throughout its life, but not light. It had also not gone yellow. Another reason to remove the label was that I would not expect it to survive two weeks in a strong oxidising solution containing sodium percarbonate/hydrogen peroxide.
After the two week soak, the case looked just as it did after the initial treatment in 2011. Perhaps slightly whiter as the treatment done in 2011 was only for a few hours. Comparing the outside surface of the case to the area where the label was, as well as the inside of the case, both areas never exposed to light in the entire time the VIC existed, reveals that the outside exposed area is very slightly lighter and chalkier than the protected areas.
A closeup of the label area shows how it is slightly darker and does not exhibit the slight chalkiness of the surrounding areas.
The enhanced view below shows this a bit better.

I also noted that areas on the case that had slight scuffs or abrasions also looked lighter and had a chalky effect compared to the surrounding areas after Retrobriting.
I had also tried the two week soak on an Amiga case that was cheese yellow. I got similar results to the VIC. Only the surfaces that had been exposed to light (and yellowed) showed any chalkiness or unevenness.
In conclusion, areas that had never been exposed to light, and never been yellowed do not exhibit visible changes or damage, even after a two week Retrobrite soak.
Damaged areas may exhibit chalkiness, but even after two weeks, the Retrobrited case looked far better than if it wasn't done.
In my opinion, assuming the retrobrite process was done evenly, and not by painting and cling wrap, then Retrobrite in itself does not damage plastic, the plastic is already damaged as part of the yellowing process, and the damage becomes visible once the yellowing is removed. The chalkiness is more noticeable on plastics that were originally darker, but even after two weeks, they still looked much better after Retrobrite compared to being cheese yellow.
Of course I'm not suggesting that items be soaked for two weeks, after all, at least in my case, after the first day there was no yellowing left to remove!