Wondering if there are IUPAC rules or conventions about structure orientation.
There are recommendations, which aren't really rules per se (in fact even the IUPAC nomenclature books are all technically recommendations). For "how to draw structures", the appropriate recommendations are:
Brecher, J. Graphical representation standards for chemical structure diagrams (IUPAC Recommendations 2008). Pure Appl. Chem. 2008, 80 (2), 277–410. DOI: 10.1351/pac200880020277.
which is free to read online. The relevant section is GR-3.4.2, which states (emphasis mine)
For ring systems containing at least two fused rings, that system should be presented in its preferred orientation. For most practical purposes, it is sufficient to orient the ring system so that:
- The fused ring system is oriented more or less horizontally. A symmetry axis (or plane) should be vertical or horizontal.
- If the system contains multiple “rows” of rings, any “extra” rings should be positioned so that they are toward the right and the top of the ring system.
- Larger rings should be toward the left and smaller rings toward the right.
- Heteroatoms should be toward the right and, with slight preference, toward the bottom.
So, generally, according to these specifications, caffeine would usually be drawn with both rings lined up "in a row". Of course, nothing is stopping you from drawing it differently if you wanted to do so. However, it would almost certainly look out of place in formal scientific writing.
[NB: Some people like to think that IUPAC are being a pain in the back when they make such recommendations. The truth is that by far most of these recommendations already exist as "common sense" or "conventions" to chemists, and they already obey most of them subconsciously. Furthermore, these recommendations aren't enforced by anybody, anywhere, so deviations are always possible (and do occur regularly).]