Let's get rid of the words fact and theory for a brief paragraph. I believe that you can use 'repeatable methods' (observation by experiment) to demonstrate how penicillin works. Our understanding of this is falsifiable, which means it can be shown to be wrong, but it hasn't been. Our understanding of it is comprehensive.
Does this make it 'truthful' or 'factual'? Well, yes and no. YES, in the sense that everything we know supports it and practically speaking it provides the result we expect time and time again. There appears to be very little that we do not understand about it. On the other hand, there will always be a NO, in the sense that our methodology might be wrong and we take assumptions for granted. In microbiology, this includes the very lowest level, the atomic level, as well as higher cellular levels.
For example, when RNA virus immune crops were first planed and grown for harvest in the US, the world had a very clear idea on how this technology worked. The transformed plants would produce viral coat protein in their own tissue and as this would be in such high abundance in each cell, the invading viruses could not unwrap - the plant protein would replace the viral protein and stop viral replication occurring. By your own logic, this would be factual as it could be demonstrated by experiment. Or so we thought.
Over a decade past and in the late 90s, scientists found out this was utter crap. It was shown that this in fact did not happen and nobody had a clue how it was working. All they knew was that it worked. The theory was wrong. It was later discovered that by causing the plants to replicate viral DNA, this was pre-empting the plant's RNA dicer system to shred up the viral DNA before it could replicate - an immunity of sorts. Very clever. But this demonstrates why we do not say theory is factual.
The point it, as there is always the no element, it would be downright foolish to regard scientific theory as providing 'truth'. That would be reading in-between lines to find things that need not be read. Observations and facts are taken to form theories, but those observations themselves can be based on other theories (how the beta-lactam ring breaks down at an atomic level, to use penicillin as an example).
The correct approach to take is to avoid using the word 'truth' when talking about broad scientific subjects altogether, as generally it's completely unsuitable. Whilst things may appear 'factual' in a practical sense, in the context of philosophical conversation broader scientific observations rarely are 'factual' as they are largely based on much theoretical workings. This is not to say that they are not likely to be correct, far from it! Is it simply that science is limited in a philosophical scope - if you are from a scientific background, as you state you are, then you should acknowledge this.
Rather than say 'our understanding of penicillin is factual', it is much better to say 'we have a fully comprehensive understanding of penicillin based on our current observations and our current understandings of the universe'. I would allow you to say the former in a strictly scientific conversation, but not a philosophical one such as this one, because it's grossly inappropriate.
