Sizing die compression

I was sizing some Lyman 358495 .38 wadcutters and noticed that the boolets are compressed and the lube grooves are much narrower than before sizing. They drop from my mold at .360x361" and the sizing die is .358" I've also noticed this occurring with a .45 Colt bullet in a .454 die - narrower lube groove. I'm using range scrap of unknown hardness, Lyman sizing dies in an RCBS Lubrisizer II. I saw a recommendation in another thread to tumble lube the boolets before the sizing/lube step but that seems like a real pain so I haven't tried it. I haven't seen any leading as a result of less lube, but I'd rather have the grooves stay as cast. How have others dealt with this problem?

Anytime you reduce the size of a boolit the lead will flow into a void first (grease grooves) which will reduce the size of the grease grooves. Which reduces the amount of grease that the grease grooves can hold.
Standard belief is: Boolit should be sized down no more than 0.002" from cast diameter for best accuracy.

Hickory is correct. I will add that having all the lube grooves full of lube is not needed in the vast majority of 38 wadcutter uses.
Over the many years of full wadcutter usage in 38 special, it has been found and repeatedly verified that lubing only the bottom groove gives the best accuracy for that boolit in most handguns.
Another thing to consider is that the boolit will get sized down when going through the cylinder throats and into the barrel anyway. Best too have the sizing done beforehand so it is done evenly and concentric.
The extra area on the drive bands may grip the rifling in the barrel better than thin ones would.

Testing how much a particular bullet design can be sized before accuracy deteriorates is going to be difficult and require a lot of shots fired. Even if someone does it and establishes a supportable conclusion, the results will only apply to one bullet mold and one gun.

I've read several times that sizing can be detrimental to accuracy (Cast Bullets For Beginner and Expert, by Brennan). For my shooting, .003" at most for my handgun bullets does not show accuracy loss, but I'm not known as a Marksman.

As mentioned above, when sizing bullets, the excess lead is not removed, just moved and it's gotta go somewhere. The crimp groove, lube grooves and the bullet's base are places that it shows up...

Two ways to solve this:

1.) size boolits .360
2.) get a mould that drops them .358

Thanks for your comments. Since narrower lube grooves aren't presenting a problem, I'll live with it.

I cast the #358495 out of COWW or 20/1 for about 10yrs, until I realized I COULDN'T cast better then I could buy swaged HBWC's.
Lubed only the bottom 2 grooves, sized .358, never noticed the lube grooves being "squashed".
My DAD used to cast the old LYMAN #358395 out of 20/1 and then Pan Lube them. He could shot perfect round one hole groups at 25yds out of his COLT OFFICER'S MODEL MATCH. But then He could bust Jackrabbits at fifty yards with a RUGER 3 screw Blackhawk. Or a COLT OFFICIAL POLICE with .38-44 loads.

Just never noticed driving bands being "squashed" at Lubing/Sizing. Have to check that out next time I cast some LEE .44 WC's.