Quite nice. I've always used Rodinal at 1:50. Thought I read that at 1:25 it was more grainy. My sense is that 1:50 is by far the most common way Rodinal is used....but I could be wrong.
I had it the other way round: Due to the acutance effect, it also very slightly accentuates the grain at 1:50. It is better from acutance perspective to use it at 1:50, and to agitate not more than 2 or three gentle inversions once a minute (or less often). By using less agitation, one can pull up the mid-curve, but one must remember to increase the total developing time in that case. At 1:25, Rodinal is somewhat less special in terms of acutance, and grain is not much different between the two dilutions. So unless you need shorter dev times, it is better to use 1:50. Some films have really long dev times at 1:50, but they are the exception. FP4+ is one of the culprits, as is Kentmere 400.
So, what do you think you are doing to minimize grain? I'm apartment living right now and decided I would use Rodinal for convenience. I immediately noticed the grain with 35mm trix. Does lower temperature help minimize it?
Yes, developing at 16 degC will reduce the grainy appearance, but it cannot be eliminated for a film such as Tri-X, HP5+ or similar. You either like the grain or not, and if you hate it, then Rodinal is probably not your developer and Tri-X is probably not your film. It is worth saying that the higher the temperature, the worse the grain issue becomes. So if you like the Tri-X in Rodinal look, but are slightly grain averse, keep the temperature down as much as you can. If you want less grain, then seriously consider TMax 400 and develop it in D76 or Xtol. That will give better results from a grain perspective.
The number of grains and their physical size of course do not necessarily change, but their appearance and the way they develop into one another do change. So one can get fine, sharp grain or blotchy or lumpy grain. The fine sharp type is preferrable to me, and is what you will get if you use Rodinal and tame the temperature (20C/68F and below). I always think of Rodinal as an honest developer - it shows the film as it is. No solvent shenanigans or anything like that. However, there is the trick of diluting the Rodinal 1:50 or 1:25 into a 100g/L sodium sulphite solution. That will give a solvent effect which will reduce the grain appearance similar to D76 at stock strength. It is a much more expensive way to use Rodinal, so is not often done.
EDIT: One more thing - if you don't like grain then don't overdevelop the film. You will have the best grain result with a negative that is as thin as you can afford to have it while still getting the full scale that you want to print. If you must, then increase the EI and trim back the developing time. As with most things, you'll have to do your own testing here and not solely rely on what others tell you. This is one of the reasons why 35 mm is developed thinner than 4x5, as on the latter the grain is hardly ever an issue and as a denser negative can accommodate a wider tonal scale it is preferable to err on that side.