You will be bound to a torque limit by the displacement of the engine and the fuel type used.
You can not make more torque.
However, you can shift that torque around to different locations on the rpm band. Shift it low, you get broad daily functionality but relatively low hp numbers. Shift it high, you get a more sport geared power band and high hp numbers. Torque is the instantaneous force the engine applies to the road at any given moment. Horsepower is the application of this torque over time, the energy expenditure of the engine. HP = energy out.
The normal goal of na builds is to first optimize the system. You raise the torque to its limit. You also broaden the torque band over more of the power band to make the car more broadly usable. Beyond the slight gains from stock, you will never see any more raw torque gains from the car. The goal from then on is to keep shifting the torque band higher in the rpm range and bumping up HP. Depending on the dyno, you will typically see between 150-170 ft-lb. from the Subaru engine. In stock form or anywhere near stock form, this torque band is centered around 3500 rpm and rolls off in both directions. Some mods will shift it a little higher. Some mods will broaden the torque band. You will see most Subaru engines not make very high HP numbers simply because the torque band is geared low. You want to make big hp numbers, you have to gear the engine for high rpm operation. A lot of this comes from the designs of the intake manifold, heads, cams, and exhaust manifold. These parts will largely define how air flows through the engine and where peak efficiency will occur. Keep moving the torque band up and you will keep making more HP. I mean, if you can keep 150 wheel ft-lb at 8000 rpm, you'll make 230 wheel HP. If you can get it to 9000 rpm, you're at 260 whp. You get the idea. You are always bound by the size of the engine, but rpm can net you major HP output...if you are willing to build an engine for it.
The downside is that the efficiency range is really only good around 4000 rpm or so. If you have an engine that will actually wind out to 9000 rpm and have the efficiency range between 5000 and 9000 rpm, it'll run like crap below 5000 rpm. The car will idle poorly, have more cleaniless (poor burning, high smog), idle poorly (may need to idle up at 1500 rpm to keep it from dying). There is a certain something you give up when actually building a race engine.
The other way to make power is to change the fuel. Put more energy into the combustion chamber, and you will get more power out given the same displacement. It's why nitrous is used. It's why people step to E85 or race fuel. There are still limits to what the engine can support in terms of load and heat, so you can really only go so far safely. Maybe you'll opt to seriously build up the engine.
I'll leave on a final note. There is a way to make a car faster for free. Two words = weight reduction. Lower static mass and lower rotating mass. Make the output energy from the engine go further. Reduce waste. A 3300 lb. car with 300hp is the same as a 2200 lb. car with 200hp. Rotating mass is another big factor. For example, there's a lot of energy wasted spinning up the stock flywheel, and a light weight one uses significantly less energy to spin up. On smaller levels, this also includes pulleys, wheels/tires, driveshafts, etc. Many times you're removing both static and rotational mass. On a quarter mile drag race, a light weight flywheel will make it so the car will use about 2.5% less overall energy from start to finish. This is quite dramatic just putzing around town too. Reving up in 1st gear alone from idle to redline is a solid 7% reduction with a lightweight flywheel. Weight plays a major roll in the whole scheme of making a car fast.