Everybody at the track is correct.
What you are feeling is driver feedback through the steering.
When you have -1° camber but the car leans over 3-5° the front end washes out mid corner [ but it will turn in OK because the car isn't leaning over yet]
I'll try and explain it easier .
Go going back to newton's laws of equal & opposite reactions. Moving mass will continue in a straight line until it is met with an equal and opposite force.
So Weight needs to overcome by Steering [traction]
So at the LIMIT of traction the heaviest end want's to continue in a straight line [understeer in a front heavy car, or oversteer in a rear heavy car]
Most vehicles have stiffer suspension to support more weight, hence the wheel rate on the front suspension is stiffer than the rear.
Now here is the Trick.
When the front end of the car bodyrolls 5°, the rear end of the car also bodyrolls 5° [unless the chassis is soft]
So if you stiffen up the rear end so at the same G's of weight transfer it will only bodyroll eg: 4° , the rear end will support more of the total weight transfer.
In simple terms you are tricking the TYRES into "thinking the car has more rear weight"
So it will over load the rear tyres Before the front tyres reach the limit.
[the opposite to what you have now]
A lot of people misunderstand this concept because they only think about roll centres when both front and rear RC's are actually a roll axis [the whole chassis rolls together the same amount along this axis]
The Stiffest end will always slide out first.
If you cant get an adjustable rear bar, install stiffer rear springs.
---------------------------------------------------------------------------
Kerry is absolutely correct so far as it goes, and if you read books like "Tune To Win" by Carrol Smith, it all makes sense.
However.
The fly in the ointment here is that all of the above assumes proper contact-patch management. Remember that every control input to the car translates THROUGH the contact patch. If you alter the contact patch, you also alter the behaviour of the chassis as well. Throw some race rubber (hi-grip) up front, and put bicycle tires in the back, and you WILL have an oversteer machine, without altering the roll centers, roll couple, or wheel rates.
We drive MacPhereson strut cars, and those are a "special case" exception-to-the-rule. When you load a McStrut car up in a corner (entry or mid-corner, pure roll analysis), you will naturally compress the outside front suspension. With a McStrut, the more you compress the suspension, the more POSITIVE camber you induce to the static setting, and it increases on a rising curve, not linear. With a static negative camber, let's say stock-ish at around 1.0*, it doesn't take much loading to have the tire "standing up" vertically, maximizing the contact patch.
Easily achievable with a stock-sized all-season tire. Throw on some sticky street tires? There will be enough load transfer onto the outside front to compress it to the point where the static -1* isn't enough to compensate, and you will wind up with POSITIVE camber at the wheel. That means you're riding around on the outside edge of the tire, not the full tread face. Reduced contact patch means reduced tractive capacity, and thus that corner looses traction first. Note that the rear (solid rear axle) doesn't exhibit nearly that much of a negative effect, as it's anchored by a bloody great chunk of iron, and the only real camber loss is found in tire carcass shift.
So, flying in the face of most tuning theory, if you STIFFEN the FRONT wheel rate in roll (either through stiffer springs or a heavier bar, or both), you will reduce the amount that the outside suspension compresses, allowing for a maximized contact patch without excessive static negative camber. If you have a 9" wide rear contact patch, and a 5" wide front contact patch, the front is going to give up first. You can, as Kerry suggests, dial in -3* of static negative camber, which will help dramatically at the track, as it allows SO much more camber-gain before the tire goes over-square. It will also cut into your braking ability and wear the inside edge of the tire fairly badly at all times that it's not cornering. On the street? It'll cut the life of your tires by 75%.
If you're trying to visualize the front suspension (front view), you have what essentially amounts to a triangle. The FLCA pivot points are fixed, as is the strut mount at the top. The apex with the tire is free to pivot, as the strut can compress and expand, and the ball joint allows change of angle. Of note, here, is that the spindle (and thus wheel) is fixed relative to the lower portion of the strut. Static, the FLCA is roughly flat relative to the track. As the suspension begins to compress, though, the FLCA pivots upwards at the ball joint, the strut compresses, and the entire tire/wheel/upright moves upward and inward. Simple logic will tell you that as it moves inward, it will gain positive camber since the top pivot point (strut mount) is fixed. Thus you gain positive camber. To combat this, you have four main options.
1) Reduce the load transferred to the suspension. In other words, slow down and don't corner so fast. (Not exactly what we want to do!)
2) Add static negative camber to "preload" the angle so that in the corner it comes back to near zero, which maximizes the contact patch area. It also burns the inside of the tire off in short order and will also reduce the braking capability somewhat.
3) Reduce the tractive capability of the OTHER end of the car (heavy rear bar/spring) to match. It'll be slow, but at least it'll be balanced... (BOO!!)
4) Increase the wheel rate to reduce the amount of motion, and add a bit of static negative camber. Heavier spring rates, or higher bar rates or both will accomplish this, at the expense of ride quality. This effectively "locks out" the front suspension in roll, but still allows some movement for longitudinal load transfer (accel/braking).
FWIW, I'm set up with 750lb front springs, the Steeda bar on full-stiff (Just about the same as the Strano bar), run either Hoosier R or Pirelli DH race rubber and have a -2.7* camber spec. My rear bar (Strano) is on full soft, and I run 300lb springs in back.
The "youtube tags" didn't seem to work for me, so here's the URL:
https://youtu.be/LmNb2Vq-GBs
It's a pair of 80% pace laps, in a track-day environment with full point-by passing rules in effect. Essentially a Sunday drive, but still good enough for around 1.5G in the corners, even on a bumpy track. About 7 seconds off "race pace," no sliding, no drama. Had to behave, as I was instructing that day... ;-)
The camber-gain issue is why "proper" race cars run SLA designs.