Obviously this is to the best of my knowledge, it's not etched in stone, and I typed it up in 3 parts due to other 'real world' stuff going on, so if a paragraph or sentence seems unfinished or something, let me know I'll try and fix it!
Ok windage tray in general - to keep oil from frothing or splashing up and hitting the crank. At 8000rpm (for example, the outer edge of the crank counterweights and the rod big ends are doing some serious speeds, and hitting even small volume of oil will be a serious impact (costing power for the most part) . So the windage tray basically allows oil to drain down back into the sump, and discourages it splashing back up over bumps or around corners.. On an exceptionally smooth track, it'd be less of a benefit, but on any street circuit (or dare I say any race track in Aus) it's worth the (relatively small) effort to add one in. It's even possible that without oil physically touching the crank from the sump, that the sheer velocity of the crank will create enough airflow/velocity to start picking up some of the oil - a little bit like how tornados out at sea can pick up columns of water (1 or 2 sydney to hobart yacht races back I remember some vivid footage off one of the boats of that very phenomenon)
On a related note, you have hinged doors in the floor of a winged sump, it'll allow oil to flow toward the centre around corners, but will resist it flowing out and into the wing (i.e. hard left corner, the right wing door will stay shut discouraging the oil running all the way out and up the right wing/wall, but the left door will open allowing that oil to flow to the centre of the sump and keeping that level higher preventing the oil pickup seeing air. And vice versa. i'll do a diagram in ms paint if that is difficult to visualise (and I readily concede that in this case the written word is a poor second place to a diagram)
Crank-scrapers - simple enough concept - although oil is generally flung off the rod big end area, some of it will tend to 'stick' to the surface of the crank. This happens to varying degrees (and generally the rougher the crank/counterweight surfaces the more they will grab) on varying engines, and this too is basically extra mass that has to be spun around, so it can cost power. The scraper is nothing more complicated than a sheet of metal on one side of the block (you can attach via sump bolts, weld into the sump, via the main caps or combos thereof) that is cut to reach within say 1mm (or closer) to touching the spinning crank, so it's shape would be a weird outline to follow the crank shape at that point. (I'll do a diagram). It sits that close, and any excess oil that is clinging to the crank beyond some token 1mm (or less) will hit the crank scraper, and be 'wiped' or scraped off. You'd want that to be able to then drain into the sump with minimum fuss, so (*due to the direction the engines rotate) you'd have the scraper sitting on the driver's side of the block (or the passengers side for those of you with LHD cars!!) - that way the oil is scraped off the crank there and can drop straight into the sump. If you had the scraper on the other side it'd scrape if off, but end up on top of the scraper, and then it'd want to keep hitting the crank again as it tried to flow around the edge of the scraper and then down into the sump..
For all of these things, the longer stroke engines (a14 and a15) have (for any given rpm) much higher outside circumference speeds, and reach further out toward the sides of hte block/sump and closer to the oil level in the sump, so the condition or prevalence is magnified vs an a12
http://jmac.performanceforums.com/crank%20scraper.jpgI've used (and enlarged) a picture from - 'crank-scrapers.com' - which shows a good one. According to the site they make them for datsuns and an L series, but it could be diy-ed by anyone fairly easily. If they object to me using the image, I'll remove it immediately, I have only saved and uploaded it to my website so that their website isn't slowed down each time this forum post is accessed, if I had just linked to their pic on their site. You'll note on their one, they have a shim on the other side of the engine to even up the spacing for sump gasket sealing purposes. The picture I linked is not their datsun item, as that picture wasn't as clear for illustrating the concept imho.
Power gains from a crank scraper probably on the order of 3-4bhp, doubtful more than 5bhp, on a largish A series or smaller capacity out of the L family.That doesn't sound like much, but it's perhaps a gain of 2-3% on a race oriented engine, and for about $5 it's the cheapest horsepower you'll ever make on an NA performance engine.
On a very very similar vein, one of those low buck (just more labour intensive) things you can do - well I mentioned smooth/polished crank counterweight surfaces - well you can do the same thing to the head, and the block, in any places where oil drains back to the sump. A shiny slick surface will speed up the oil drain back and massively reduce the oil's tendency to leave a film clinging to a rougher surface. Which means the oil is back in the sump sooner, so there's less chance of the pickup sucking air, it all helps. It also helps prevent too much oil lingering up in the rocker cover, which would otherwise be more likely to leak out the r/c gasket, and or overwhelm the valve guides (on racing minis without enough breathers, believe it or not after a 1 minute flat out thrash, if you stop quick enough and pop the bonnet and remove the oil cap, the rocker cover can be full near to the top of oil. Now the dattos aren't anywhere near as bad, but it all adds up, and there's no drawbacks to this work).
I have no data to back this up, but I've always wondered as to whether you could go too far on a crank scraper on a flat tappet cam in block type engine, as it would potentially reduce the oiling to the cam lobes. On a roller cam, likely not an issue, they get by with less oil anyway. On a racer - heck, you only have to get it through x number of seasons between rebuilds - nothing like street mileage! you'd probably have to go closer than 0.5mm to be an issue like that, but as I mentioned, I haven't got any real world data on a crank scraper on a street driven engine that requires 150,000km+ longevity.
----
very short summary - on wet sumped engines (and to a degree on dry) oil control is a massive issue for circuit racers, it's either hitting the crank on drainback, draining back too slow, frothing up from the proximity of the high speed/rpm crank/rod/counterweights or sloshing up during a hard corner. Anything and everything you can do to keep it in the sump in general means the engine is all that much safer from oil starvation related damage. In addition to that (i.e. longevity/protection advantages), oil clinging to the crank, or hitting it as it's dripping down or splashing up costs power. Not a million bhp, but definitely some. So the windage tray and crank scraper work to minimise this issue from the various ways it can come about. In a tightly contested circuit race series, 3-5bhp on a moderate capacity 4 cylinder is a hell of a gain.
----
Dry sumped engines avoid the oil splashing up etc because the oil isn't stored in the sump, but in an auxilliary reservoir. there are pickups in the 'sump' to suck the oil that drains back to the bottom, and shunt it to the external reservoir where it's pumped back into the main oil circuitry. With all that oil 'gone' from the sump, less is frothed up, and so forth, so the dry sump (aside from it being able to keep oil pressure up no matter how many Gs you corner at since the pump draws oil from the bottom of a funnel shaped reservoir and can never run dry like a wet sumped oil pickup) is basically the 'ultimate' implementation of the concept.
On top of that, with less oil splashing around (dry sump setups) you can run lower tension rings, as there won't be nearly as much oil to have to keep scraping down the bores. That provides a bit more power/less friction. Add on top of that (and it all works in with one another) that with less chance of oil contamination past the rings, they can safely run more compression without inviting detonation. On a similar vein, they can run vacuum pumps to get rid of crankcase pressure, which then 'pulls' oil that would otherwise get up past the rings, and it's even more the case, less contamination etc etc etc. Theoretically you could run a vacuum pump on a very worn engine, and prevent as much oil burning, BUT you'd generally want to avoid it on at least most fresh street engines, as they (with more ring tension - the oil rings of course, but less widely known is the fact that the second compression ring actually mostly assists with oil control, despite its name) will end up with insufficient bore lubrication potentially, and it could shorten their life. Horses for courses and all that.