Tin flash-oxidizes instantly on the surface of molten lead-based alloy. Tin oxide is much more thin, light, weak, and flexible than lead oxide, therefore it allows the molten, UNoxidized metal beneath to flow without the effects of its own surface tension. Think lava flowing down the side of a volcanic atoll, the hardened crust impedes the flow and must continuously crack and break apart to allow the molten lava beneath to flow. Same thing with a pot of alloy, but on a microscopic level. The properties of tin allow the metal to effectively flow more smoothly and conform to finer contours as Rick said.
The other property tin adds to lead is it alters and breaks up the long, linear planes of the lead crystal dendrites that form as lead solidifies and cools. In a lead/tin/antimony mix, tin and antimony will bond to each other up to about a 1:1 ratio, effectively making a new element in the mix called SbSn. Sb/Sn brings some very unique reinforcing properties to the lead dendrite matrix, making the alloy even more strong and "hard". High Sb with low Sn make an alloy which has distinct, long shear planes but is strong in cross, section, akin to a coarse-grained wood like yellow pine. It will split easily, but not bend easily. If more antimony than tin is present, the alloy will consist of mostly lead intermixed with Sb, and interlaced with Sb/Sn. If more tin than antimony is present, the alloy will consist mostly of lead intermixed with Sb/Sn and a lot of free tin "nodules", which can be problematic and should be avoided, so keep your tin exactly at or less than the amount of antimony in the mix. Ternary eutectic mixes such as Lyman #2 and some of the pseudo-eutectic blends like Taracorp Magnum and some others have a very small temperature window within which the "mush phase" occurs when freezing, so there is less precipitation hardening going to take place after cooling due to the dendrites being captured in a moment in time. With less eutectic alloys, the ability to heat treat and for long-term precipitation hardening to occur is greatly increased. Lead/tin binary alloys tend to be what they are, have no significant ability to heat treat, and tend to be stable strength-wise from the day you cast them until years later. This can be an advantage, but it can also be a limitation because the short, fine dendrites of the mix give the alloy a non-directional, mushy characteristic, like silly putty. Lead/tin is very malleable, but not as ductile as alloys with antimony in them because of the nature of the crystal matrix antimony makes within the structure of the metal. The structure of the antimony/lead matrix is also why antimony can add so much more toughness and hardness to lead than can tin alone.
All of these different percentages of Pb, Sb, and Sn or just Pb and Sn have different properties which can work for you, or against you in your bullet-launching efforts. Understanding what each constituent does or doesn't do, and how the sum total of your particular mix will react to the particular choice of powder lit behind it, is a big part of getting good groups. Sometimes, there is a broad range of different combinations of alloy and powder that will work well with a given bullet shape and throat shape, sometimes there are very few or maybe even just one, if you can find it. Determining what to use starts with the chamber and throat of the rifle and the intended ballistics of the load. Then the bullet shape is selected which one anticipates will cooperate to those ends, together or sometimes after the alloy of choice is made. Alloy is primarily a choice of propellant type, velocity and range, and terminal considerations. If you aren't concerned with anything except killing paper, steel, or other targets of amusement, your alloy and velocity windows automatically become larger than someone who, for example, needs a minimum of 2K fps muzzle velocity, an alloy that will offer good terminal ballistics at 16-1800 fps when it reaches the target, and reliable 1.5 MOA maximum dispersion out to 200 yards.
When casting, Rick pretty much described what I do a lot of the time. About 100°F above the full-liquidus point of the alloy, plus some if trying to get more heat into the mould beyond that of the fastest casting pace I can comfortably maintain. Bullet fillout and bullet quality are almost entirely a product of consistent and correct mould temperature, which of course is a function of mostly casting pace and to a much lesser degree a function of the temperature of the metal being poured. .22 bullets in a set of big, brass blocks will drive you nuts trying to keep hot with a bottom-pour pot. Most binary alloys at 20:1 or more rich with tin practically fill the blocks on their own just past the liquid point of temperature, so there's no need for an extremely hot mould OR alloy, in fact sometimes 16:1 fills TOO well and there is a constant fight with vent line whiskers, clogs, and little whisker remains sticking to the block faces. Also, binary lead/tin alloy loves to "tin" shiny new mould blocks, after all it is basically solder.
