Many people overlook the fact the 86x86 is Honda's version of a pure square racing engine for endurance racing and road course racing.
BTCC, STW, DTM and many other racing scenes used square near bore-stroke-ratios, which is not bad up to around 9000 rpm as limit for 2-liter-I4-engines. Opel/Vauxhall had a square design in the C20XE, the Renault F7R 2-Liter engine was under square with their 82.7x90 dimensions, the Audi engine EA113 was also under square with 85x88 (at a 88 mm bore pitch!), the STW engine of BMW, called S42, had 86.5x85 based on the M42B18 (84x81). Renault won more then 50 % of the races of 1997 I think with this under square design. But remember these engines were limited to 8500 rpm per rule book and made end of the 90'ies all around 305-320 hp. None of them made more then 26 m/s mean piston speed or more then 40 m/s peak piston speed. These are quite well experienced values for todays of the shelf quality material in the lower comp piston area.
The F7R engine or the F4R engine of Renault was a real beasty boy, 200
[email protected] and 215
[email protected] rpm in series. This is quite a step up of the K20A2 engine and competes directly with the JDM version with 225
[email protected] rpm and 215
[email protected] rpm. The rod-stroke-ratio was just the same with the K-series, despite the 90 mm stroke because of the bigger block height. That this worked quite well even with the tiny 32.3 mm OE intake valve (83.7 mm bore!) size. A different approach as the K-series does have it with the 35 mm valve. Finally I don't know what sizes they utilized in the race version, e.g. if bore was increased to favour a shorter stroke and bigger bore for more air, but the basics are different to square design and was great success.
I've calculated the piston pin and rod tensions as well as the forces around the crank drive as well as normal force at the liner-piston-skirt for my DAMPFHAMMER HD engine to analyse how these went on when engine speed is pushed further to 10,000 rpm. The F20C1 setup is quite racy with the over square design, the long rod approach and 2 mm lower stroke helps allot to reduce side force and piston speeds. It's understandable that many racers go this route with the 2-Liter-10,000 rpm area. But to be honest, for the 330 hp approach it isn't necessary. For my 360 hp approach it would help, but it means a huge invest. The R/S-ratio of 1.82 won't be reached with the 84 mm stroke in a 212 mm block height, would be 1.67 to 1.71, depending on compression height, that means allot on the alternation of load side as well on the friction side. The way to go to 89x80 would be attractive on the alternation of load side but would loose much of the mid range torque. Joe McCarthy's 90x78.5 approach is way above the valvetrain capability, which I would limit to 10,500 rpm. The head for that engine need to be borne, as this thingy need very tight ports to come to live. I am convinced Joe focused much on the side load and mean piston speed, but power wise this thing is limited. It made with huge valves a bit over 300 hp at flywheel at around 9700 rpm, but that level can be achieved well around 8000 rpm with a stock near head. He designed it for endurance and the 5 miles on the salt. It was a safe approach. Today we know 10,000 rpm based on a 86 mm crank for circuit are not the limit when built and designed correctly. But keep in mind this are 28.7 m/s (~5600 ft/min) mean piston speed, that's an enormous load on pin, rod and bolt's, you need to be very very light in parts to stand this.
The NASCAR cup engines are in a similar but bit lower range at 10,000 rpm because of their bigger rod-stroke ratio. Now as they as speed limited the play of the game get's really nasty regarding rod-stroke-valve size design, fighting for the best wide torque bandwidth to get faster into peak power out of the corner, designing resolution is 50 rpm. These guys do a great job for the V8 system and brought the valve spring development over the years to an point were we are today. The US market has some of the best valve spring companies, just because NASCAR always screw up power with the valve spring engine speed capability.
Finally, the by the OP aimed 350 hp at flywheel are quite interesting, with ITB's very challenging. I would even say, none in the market would reach it with pump fuel just with ITB's only on intake side. The top dogs at drag racing where rod-stroke-ratio's down to 1.43 are used and alcohol fuels are allowed reach out for 180-190 hp/Liter, which would be 370 hp. But that is not transferable, as the 86 mm bore is used for the head already and it's huge 35 mm double valves. You would need to do a 80x99 engine in a K20 block and go down to a 1.4 rod ratio, which shrouds the valve hugely. So that concept of low rod-stroke-empowerment of alternation of load doesn't work here. The concept would need around 10,800 rpm to get into that range where those huge ports are on spot and then you need a let's say 500 rpm safety range to redline, which would be at 11,300 rpm. These piston rings would sing you a song about oscillation because of the almost 8000 g piston peak acceleration, which is still lower at a 106 to 108 mm stroke engine at 9500 rpm, where the 180-190 hp/Liter came from. So that concept also doesn't work. The 2-Liter engine is the real challenger for peak power competition as power comes from different sources as the typical top dog drag race engine from JBM, 4Piston, DragCartel, ... came from.
The approach stated by
@Henry Collingwood with the long rod approach (up to around 165 mm for a 3-ring system) doesn't reduce the mean piston speed as stroke is still the same. It will reduce side load and peak piston acceleration, that's all. But it increases rod weight which compensate some of the rpm's. Bolt-wise no single advantage. Pin-wise an advantage, but not the one one may wish. Side-load wise an advantage, but does it count if VE is lowered? This engine just need smaller ports to come to alive at same engine speed bandwidth and therefore it is more limited more peaky, nothing for a Miata RWD regarding control of the rear. I am convinced the 370 hp are possible, but you would invest in parts development as well as combustion process design as these are the weak links for that concept. 300 whp on a RWD none have achieved just by an 2-Liter 4-banger allmotor. You would need at least 5 cylinders, better 6 or 8, to increase engine speed, increase valve area and need to use a more revvy valvetrain (mechanical tappet system) to get 185 hp/Liter on a 2-Liter basis. On a I4 basis you are limited on many areas to reach out this level. I know concepts doing 230 hp/Liter on a simple ITB system out of 2-Liter displacement in NA mode on alcohol fuel, but this is a V8 system based on the Suzuki Hayabusa 1350 ccm heads. Short stroke, huge valve area, a Hayabusa near rod-ratio, and enough play to rev up to 15,000 rpm, does make 460
[email protected],600 rpm and runs the 5 miles on the salt. This concept is indeed reliable, a 370 hp 2-Liter K-series is it not in any case on an allmotor alcohol fuel basis.
What I would recommend is a 1600 ccm concept based on a Hayabusa engine, which are well pushed to 250 hp at flywheel, based on a more reliable concept. It would be also possible based on a B16B basis, but their valvetrain sucks above 10,500 rpm. Mechanical tappet actuation is the key to rev much higher. The Hayabusa head has a nice potential and flows very well, has a much faster chamber then any B- or K-series. In the 1-Liter class I would start with e.g. with the RR1000 BMW, very nice high engine speed design. Here you find much of the V10 F1 engine BMW had learned off.
The DAMPFHAMMER HD is spec'd for the aimed goal, but is far from realization, that money pit need time, or sponsors, to be built and validated.