Troy is correct, bullet lube not only aids in keeping the bore from leading, it is also a seal.
This is not my picture, several years ago I was having a conversation with someone on another website about round vs square lube grooves. He sent me this picture.
If you look at the lube grooves you'll see that they are compressed. This is that alloy/bhn/psi load pressure thing.
The watered down version:
You hit the loud button and "BANG" the bullets off to the races!!! As the bullet goes down the bbl the pressure of the load pushes on the bullet's base. This pressure compresses the lube groove causing the lube to press outward and forward. Some of the hot gasses get by the bullet's base (bottom drive band) and push the lube forward sealing the middle/top drive bands.
That's why it's a good thing when testing cast bullets to test the same bullet/load with soft and hard alloys over a chronograph. It extremely common to find that a softer alloy will have 30fps/40fps more velocity for the same load when testing +/- 20,000psi loads. The softer alloy compresses faster/better allowing the lube to seal better making better use of the pressure of the load, more efficient.
Not all bullets are created equal, I tested these bullets along with 2 other bullets not pictured in a snubnosed 38spl testing p+ loads.
There was as much as 70fps difference between the best bullet design and the worst bullet design. While 70fps doesn't sound like much it makes a huge difference with loads in a snubnosed 38spl. We are talking 880fps vs 950fps.
This is also where coated bullets really shine. Coated bullets don't need to have the correct alloy for the pressure of the load to get the most out of that load. The coating is already everywhere it needs to be and seals everything. In head to head testing with traditional cast/lubed bullets vs coated it's common to see +/- 50fps more velocity from a coated bullet.