It’s usually the filter transmission characteristics that determines the outcome.
The additive combination of RGB filters can result in a natural looking image, however this does not necessarily depict a “true colour” image or the way we “think” the object should look based on scientific data. This topic of true colour imaging opens up a never ending debate.
For stellar objects, filters such as UVBRI are a better match than conventional RGB. In fact, V and B match a human’s green and blue sensitivity when correlating the wavelength spectral response. I is better than R as stars predominately radiate with continuum. In other words they don’t always fall in the visual emission lines. However, depending on the filter characteristics the wavelength of I can be vast, so mapping it to an R channel can result in weight problems if not managed correctly.
As the core of globulars are heavily populated/dense, there is typically a “haze”of radiating light. This is depicted with a subtle cool white tone through filters and can obscure/mask the stars past main sequence. Some filters will show this with a blue tone (which can be a weighting problem). I prefer to call it cool white (like what you get from the flourescent lighting - compared to warm).
Few astrophotographers use UVBRI filters for imaging “pretty pictures” but those that do obtain remarkable results. As an example – Stephane Guisard has produced an RVB image that displays a very appealing view of NGC5139 -
http://astrosurf.com/sguisard/Pagim/Omega-FS128.html. You can clearly see plenty of the older stars, partly through the dense cool white haze. There are other RVB images on his site.