You might look into lith printing with a suitably grainy paper as well. It introduces what the lith community calls "paper grain" but it's not really paper grain, it's more like hydroquinone radicals affecting nearby grains at points of development. Regardless of what it is, it's a very attractively grainy appearance. Be warned that you'll be truly sucked in once you try it though and it's a fairly time consuming process (5-20m per print depending on paper, dilution etc). To me though, there's nothing else that can match the results no matter how much you mess with the negative or regular prints.
For getting the grain on the negative though, keep in mind there are two types of grain visible in a print. Shadow grain is from printing close to the dmin of the film. The dmin is naturally grainy due to manufacturing process (and many sensitizers basically work as very mild fogging agents). If you're printing shadow detail very close to dmin, you'll get some of that dmin into the print. This is why thin negatives typically print and scan more grainy than pushed negatives. The other side is more like, idk, "variation grain?". Basically it comes more from the actual grain designed into the film, but the developer can affect this grain. Silver grains are variable sized, some big some small. Bigger grains are more sensitive to light and smaller grains are less sensitive. So, in a natural "normal" developer, the grain is more from variations between these sensitivities. For example, you take a picture of a grey card, but some grains are lighter and some grains are darker. Some developers are designed for "solvent type" finer grain. They don't make grain go away, but make the edges between these differences smoother and rounded off, hence giving less obvious grain even though it's very much still there. Other developers are "replating type" finer grain. They work by taking some silver from the bigger grains and depositing them on the smaller grains, thus averaging out that difference to an extent. This basically touches on "definition" and "sharpness/resolution". And of course, most fine grain developers work by incorporating both of those aspects. And then other developers work in the opposite way, to make the difference between these grains bigger, or even the "edge effect" stuff where development locally exhausts, exaggerating the borders between bright spots and dark spots. The most obvious type of developer that does this is tanning developers. I feel like these "should" be very coarse grain, but the tanning effect masks that. It builds a mask over where it develops, bringing overall density up evenly without regard to the grain and makes the difference between the grains smaller in relation, thus hiding it.
How this connects to printing contrast and grain.. Increasing contrast on the print will increase the contrast between these grains, thus, grainy appearing print. However, negative developers are much more complicated and typically don't build contrast between grains in the same way... I mean, they do to an extent, but it's not at all like a flat contrast increase. It depends on surrounding density, design of the developer, how permeable the film is, etc... and even with that, lowering the contrast in printing will decrease the contrast between the grains in a flat way. This is why some films/developers will give weirdly grainy skies when pushing, but another combo might give you smooth skies but grainy shadows. There's a lot more flexibility to controlling grain on the negative than when printing
edit: Note that bigger grains are also less stable. They can be prone to fogging or inconsistent exposure response. This is why super fast films made using mostly bigger grains (along with tons of dyes etc) are so grainy.
For your question of how to get the ultimate amount of grain on the negative though, I'd recommend something crazy like half frame or 16mm with super fast film. Pull it if you need extra shadow detail and lowered contrast, use a grainy developer like rodinal. Some of the most grainy negatives I have actually came from a mistake in processing some 3 ISO film. I boosted it to ~12 ISO by using a fogging solution, but that went wrong. It'd be impossible to print from in that case as it looked nearly black... but used more carefully it might be what you're going for.
Example (in ~6x4.5 format)
The most easily available fogging agent but hardest to work with is sodium dithionite, sold as "Iron Out" rust cleaning solution. It's not a homogenous powder and the amounts you'll need are extremely tiny. Keep it neutral or weakly acidic to slow the action, in even weak alkalis it'll act as a very potent fogging developer. Bromide seems to slow the development but not the fogging action. Note that dithionite decays into metabisulfite and using acids with metabisulfite will result in a sulfur dioxide gas release (smells like burning matches), so be careful and keep the amounts used small if doing anything with acids around it. It is not stable in any known solutions, but triethanolamine has been quoted by a patent to work as a preservative when at very low temperatures and for a few days. Always use distilled water as any iron etc introduced will cause it to quickly decay.
