Looking over your posts it's not an easy problem to diagnose in either case but I can certainly provide a little more information which might help to find the problem. In your case, I don't see anything that suggests that the pattern (either the material it's made from or the way that it's finished) is the problem. The important thing with the pattern is simply to ensure:
Made from a compatible material (ideally epoxy)
Can take the 65'C initial cure temperature
Non-porous (which is where the S120 comes in but it not essential, depending on the material used.)
Prepared with a suitable release agent (again, that can take the 65'C initial cure).
Under no circumstances use polyurethane board or polyurethane primers/finishes.
Avoid vinylester (such as Uni-Mould) because compatibility of XT135 with vinylester is - as yet - untested.
So, having (probably) ruled out your pattern you're left with the following possible causes:
Laminating technique
Bagging technique
Cure cycle
Material condition
Of the above, #1 is the most likely cause of big pin holes and #4 is probably the most likely cause of tiny pin holes and surface defects.
Laminating Technique
Laying up a prepreg mould is very different to laying up a component. You really need to be making extensive use of cut lines, joins and smaller panels of material that can move around slightly, relative to each other. If this is done correctly, it should very unlikely that you'll get a large cavity or void. An obvious one to check is that the surface layer is the right way round - blue film side/glass scrim side down on the mould surface.
Bagging Technique
Much the same as the components themselves, making a void-free mould with high quality edges and corners would need good bagging technique. Martin (Hanaldo) you'll have covered this on the course and this is basically what we're talking about here - really ensuring the bag is down tight into every corner.
Cure Cycle
In the case of the XT135 tooling system, this part should be REALLY straightforward which is why I've not listed it as a likely cause of your problems. The *initial* cure of XT135 is simply a slow, low cure at 65'C. After that cure you'll be removing it from the pattern and the bag before its post-cure. At this stage, you can see the surface finish. Because there's no ramp or dwell or other complexity, I would be very surprised if the cure temp is the problem.
Material Condition
The XT135 surface ply is much more sensitive to out-life issues than the backing. The reason for this is the glass scrim which performs best when it is only partially impregnated with resin. At room temperature the resin becomes more liquid and will start to wet out the glass scrim (we call this 'wicking'). This means that the material works at its very best when it is at its most fresh (with the least accumulated time out of the freezer. Although from a resin perspective there is still plenty of 'life' left in the resin (even after 4 weeks +) the wicking effect, particularly in warmer environments, does mean that this is one system that will perform better the fresher it is.
I know - as suppliers of the system - that we're not exactly impartial in this but I can honestly say that in general we're getting a lot of positive feedback on this product and that it continues to surprise us with just what we can do with it. As an example, you may have seen some photos on our Facebook and Instagram of the carbon fibre frame that a local composites wizz/bike builder is making. We've supplied him with materials in exchange for documenting his project but we've not done the work or really provided any more input than we would with any customer. He's now made all of the moulds for the frame (there's more than 10 individual moulds needed) and they have incredibly complicated and difficult geometry and accuracy required. They've all been made using XPREG XT135 tooling prepreg and the surface finish is completely flawless on every mould.
In terms of fixing it, if it is just surface marring then you may be lucky and it polishes out - deeper pinholes may be harder to repair.