Sorry this is long, but if you really want to understand the issues at play the subject gets a bit complex.
TB
Water weighs around 62 pounds per cubic foot (salt water and fresh water are slightly different, but 62 lbs. will be close enough for our purposes). So for every cubic foot of water we displace, we gain 62 lbs. of flotation power - less whatever we used to displace that water. If our cubic foot was air, it's virtually weightless and we gain the entire 62 lbs. of flotation power. If it was foam and that foam block weighed say 4 lbs. then we gain 58 lbs. of flotation. If the cubic foot was a 40 lb. chunk of solid wood or other material, it would only yield 22 lbs. of flotation (along with a rather heavy boat).
So if you fill the tank with foam, the amount of flotation it will yield will depend upon the density of that particular foam, with lighter, less-dense types of foam yielding slightly more flotation for a specific volume. Many foams (including some of the mix-and pour varieties) are classified by their density or weight for a given volume and you'll see some foams listed as "3 lb. foam", "4 Lb. foam", etc.
The idea of pour-in-place foam that expands to fill the cavity is often a bit better on wooden boats that it really is in practice. If water does ever get in there, it's hard to get it out and the foam can contribute to rot from trapped moisture, so the inner walls of the cavity need to be really well sealed (epoxy). The foam also generates a lot of heat and pressure as it expands and can actually stress, or even blow out, seams if you accidentally add a bit too much. Listening to the seams on your new boat creak and crack as the foam goes off is not a fun experience (been there, done that).
The final potential problem with poured foams is that they (like a lot of polyurethane stuff) aren't completely dimensionally stable and often shrink over time. There are hundreds of fiberglass canoes out there with what are called "suck lines" at their ends. Poured foam is the easiest and most common flotation foam used on glass boats and the scenario goes like this: They wall-off the end with a nice, glassed-in fiberglass bulkhead and pour the foam in through a small opening to fill the space behind the wall, which it does. Over time though, the foam often shrinks, pulling the sides of the hull in the stem area inward as it does. The bulkhead can't shrink to match it and soon you start to see the raised outline of the bulkhead coming through on the outsides of the hull. Structurally, it's generally still fine, but it can be pretty ugly when you start distorting the smooth shape of what was once a nice looking stem area. So, you're often better off making your fail-safe tank filler material out of other types of foam (peanuts or small sawn cubes in a sturdy bag, cut and stacked slabs of styrofoam, empty soda bottles, ping-pong balls, etc.) They may have small spaces between them that could flood and cost you a little bit of flotation power, but the result tends to be more stable and is often a better idea on wooden boats. It's also usually cheaper and you don't have the toxic fumes, heat and hull creaking of the pour-in foams to worry about.
How much do you need? That's a big question and a lot of people don't understand the issue very well. It's a wooden boat and wood floats, right?.... Well, kind of yes, and kind of no. Most wood does float, but in addition to keeping the parts of the boat itself that don't float from sinking, it may or may not have enough reserve buoyancy to also keep the canoe's former occupants afloat or allow them to self-rescue.
There was an alien on Star Trek once who referred to humans as "ugly bags of mostly water". In some ways, he was right. Water weighs essentially nothing in water, so when the average Joe jumps into the pond (or falls out of his canoe) the water in him is neutral and he will generally weigh somewhere in the 16-20 lb. range. Even if his on-land weight is 200 lbs. or more and he went totally limp, you could probably keep him at the surface with a piece of 20 lb. test fishing line - or - if he's awake, we could hand him a cube of styrofoam that's about 8.5 inches on all sides (a bit over 1/3 cubic foot in volume and displacing more than 20 lbs. of water) and he could stay afloat all day hanging onto the cube. Life jacket manufacturers know this and the foam in a typical life jacket displaces around 17-20 lbs of water.
So designers will shoot for enough float tank volume to support the hull and keep it afloat (factoring in whatever additional buoyancy the hull materials themselves may provide - wood, quite a bit, fiberglass, virtually none) and add a minimum of around 1/3 cubic foot of extra flotation for each potential occupant. That way, the boat won't sink and as long as the former paddlers cling to the swamped hull, they will all stay on the surface.
Here, however, is where things get tricky. You may only weigh 17 lbs. in the water, but that's only while you remain in the water. If you slither over the gunwales of your swamped canoe to try to start bailing it out - and raise part of your body out of the water in the process, that part quickly returns to it's dry-land weight. With your torso exposed above water, the weight you are exerting on the boat may instantly go from 17 lbs. to 117 lbs. or more - overpowering the reserve flotation built into the hull. Swanp a canoe in deep water, hang onto the gunwales and you can be there all day if you want. Swamp that same canoe in three feet of water, slither in, stand up and it will go straight to the bottom - because as soon as you stood up, your weight increased, big-time, overpowering the flotation offered by the hull and tanks.
So, while there isn't a set formula for how much extra flotation volume is best, if self-rescue-ability is your goal, the more tank volume you add, the better. The swamped boat will likely be easier to bail out and you won't need to stay quite as low in it as you bail. Float tanks take up interior space, so obviously you'll need to strike some sort of compromise. As to the difference between the weight and buoyancy of air-flotation vs. foam flotation, it's actually fairly minimal in the grand scheme of things and generally no more than five to ten percent or so at most. If you're just switching from air-filled tanks to foam-filled tanks, there probably is no real need to make the tanks any bigger just to make up for the weight of the foam.
