I ran a 1679 Type I engine in my `73 Bus for a couple of years, and even with a single 34PICT-3 it compared favorably to the original dual-carb Type IV of the same displacement - 88x69 makes a lot more sense for a heavy vehicle than 90x66.
When my son took over the vehicle we put in a 1995 (88x82) with 48IDA Webers, and again it made a better Bus motor than the factory 2-liter.
So, I would say that it's definitely worth it to go to a longer-stroke crank, but that does entail a lot more issues than a stock-stroke build.
With goals as conservative as you have, it may be more sensible to retain the stock stroke. I would strongly recommend that the crank & flywheel be 8-dowelled, and (ideally) dynamically balanced, but for sub-5000 RPM work there's no real need for a counterweighted crank. Nor is a lightened flywheel required or desired in a heavy vehicle.
As Lee said, you want to avoid thin cylinderwalls for longevity. Slip-in 87s and 94s are out IMO, and slip-in 88s or classic 92s are even worse. That leaves machine-88s, "late" 90.5s and "thickwall" 92s (1679, 1775, and 1835cc respectively with the stock 69mm stroke). Of those three, the 90.5s have the thinnest cylinderwalls (nominal 3.75mm, compared to the 4.15mm of stock 85.5 jugs). The 88s and 92s are ~4.5mm thick, greater than stock.
Your bus is heavier than mine, and you already have more carburetion so the best option IMO would be the 92s. Their cylinders require that the heads be opened up to the diameter required for 94s; two versions are available, one that also uses the "94" diameter for the case opening and one that fits a case machined for late 90.5s/classic 92s - the latter is the better choice since it doesn't remove as much material from the case, and the thickness of the cylinder at the bottom is not an issue.
If you think you'd like even more than an 1835, there are many more combinations & permutations to consider, each with its own set of potential complications - but for your stated goals I'd be disinclined to go there. A lot more work/expense for the right to brag of having a "stroker"
...also, where the stock heater boxes are adequate for a mild ~1850cc engine, moving up further in displacement adds expense there too.
I would go to a slightly larger-than-stock cam just to compensate for the displacement increase (assuming the same volumetric efficiency, an 1835 needs to pump as many CFM at 3455 RPM as a 1585 at 4000). But since you don't need to extend the upper RPM limit keep it conservative. Something under ~278° advertised duration, like a Bugpack 4061, Engle W-100, or Scat C-25 for example...those all work with a totally stock valvetrain below ~4500 RPM. You could use 1.4:1 ratio rockers on a stock cam too, but those will almost certainly need shorter pushrods and it's wise IMO to run HD valvesprings too - so if you're going to need to buy a new cam anyway, may as well spend the money there rather than on ratio rockers.
You won't get into any trouble port-matching out to the diameter of the stock metal intake gasket and blending in for ~½"; a mild mini-D port is also safe to do without a flow bench. Leave the guides/bosses alone, and there's no need to touch the exhausts. Stock size valves aren't badly shrouded, but it won't hurt to open the chambers up some. You're going to need to add at least a couple CCs unswept volume somehow anyway to keep the static C.R. down with the bigger displacement, and it's (theoretically) better done there than by increasing piston deck height beyond ~.060". That said, I wouldn't lose any sleep over keeping the deck height "tight" - the 88x76 motor I just built for my trike needed .220" deck to make up for the tiny chambers in the old race heads I used on it, and it runs cooler and gets better mileage than the 1585 it replaced.
)
.