[Varyon]

Going along with the "shorter but wider" trend, I'd say that for reaction drives there's probably a limit beyond which a ship can no longer be streamlined and still mount 6 full-sized drives, then this would count down. After that you'd have a range where you can mount 6 full-sized drives, and eventually hit the point where you basically have to streamline the ship the "wrong way" (streamlined from the side) to be able to mount 6 full-sized drives, which would similarly count down until it's literally impossible to fit a full-sized drive. Note "wrong way" streamlining mostly just means armor doesn't protect as well, although you can probably designate one foe (or even a group if they're clustered together) each round that suffers a -1 to hit you from the side, as you angle to present your narrowest profile in that direction.

Going off Anaraxes' values for the RF-1, which in Spaceships would have 3G (chemical rocket) for a single system, and the Saturn V (or rather its first stage, but I'll just refer to this as "Saturn V" here) apparently had ~1.5 systems (4.5G) - so each RF-1 is a full-sized system for an SM+8 vessel (Saturn V was ~3000 tons, so SM+9, and each rocket was 0.3 systems, going off Spaceships performance of Chemical Rockets; note this is largely just rough estimation, but that's probably good enough here). Approximating the Saturn V as a cylinder with a 33 ft diameter, and saying the smallest the nozzle could be to achieve the correct thrust would be a circle with a 1.7 ft diameter, we see the Saturn V had a cross-sectional area of ~855.3 ft2 while each nozzle had a cross-sectional area of ~2.27 ft2. You need 3 of them to make up a full-sized system, for ~6.81 ft2, or ~8% of the cross-sectional area of a streamlined vessel; we'll round up to 10%. That's at SM+9. If we further assume that the Saturn V was only just above the unstreamlined vs streamlined cutoff (unlikely, but let's go with it), that means if it had more than 10 systems worth of Chemical Rockets, it wouldn't be able to be streamlined (as it can't have more than 100% of its cross-sectional area as nozzle). At SM+10, diameter increases by 1.5x, roughly doubling cross-sectional area. Our old SM+9 rockets take up 5% of the ship's cross-sectional area, but we need 3x as many for the same thrust, so they actually take up 15% per full-sized system - we can only fit up to around 7 and still be streamlined. At SM+11, this becomes up to 5, then at SM+12 it becomes 3, at SM+13 it becomes 2, at SM+14 it becomes 1.5, at SM+15 it becomes 1, and so forth, following the Size and Speed/Range Table (SSR). You've then got a pretty large range where you can mount 6 full-sized systems on an unstreamlined drive (you need the same relative dimensions - swapping length and width - as a Saturn V for it to count as side-streamlined).

That's apparently with the reaction drive that requires the largest nozzle size, and also with us being a bit harsh (like saying Saturn V is the closest to unstreamlined you can get and still count as streamlined). I feel that unless you're dealing with really big ships, this is probably something you can safely ignore. If you go into enough depth to calculate what nozzle size is needed for a given reaction drive, you can compare this to a minimally-streamlined reference vessel and see how many such nozzles would "fit" if you went to 100% capacity. Thereafter, each +1 to SM is -1 SSR (so x0.7, x0.5, x0.3, etc) to the number of nozzles that can fit.