Amen!
I can't agree with the idea, previously expressed in this forum and in this thread but not worded exactly as I am about to do so, that you should go with your feelings regarding tilt safety. People "get used" to doing things that are just plain unsafe. FAMILIARITY BREEDS CONTEMPT Track width, wheel base and height of CG don't care how you feel. Subjective measures are patently unsafe when employed in issues involving safety. Recall the saying, "There are old pilots and there are bold pilots but there are few old bold pilots. I think tractorin' is a parallel.
Ask Harv what he thinks of the idea of ignoring the dive tables and staying at some depth (don't use a gauge, just subjective "FEEL") for some time (don't use a watch or botom timer, just subjective feel) and then coming to the surface however fast seems like a good idea at the time (don't use a decompression computer, or tables, or not out run the smaller bubbles, or anything). I know what they call those guys, B E N T (slang for getting the bends).
It is patently unsafe to rely on an emotionally modulated safety standard. If you are feeling great, invincible ready to slay dragons you might exceed the tilt that would otherwise set off your pucker factor alarm.
Due to liability concerns Kubota is quiet on the topic of "SAFE TILT ANGLES" as is our friendly tilt meter maker.
Empirical data is needed. Any volunteers to video themselves at ever increasing tilt angles until they fall over like the guy in the rainsuit on the trike on "Rowin and Martin's Laugh In"? I thought not. (Under 40?, ignore laugh in reference or ask someone with more grey in his beard.)
If I had the means at hand and some free time, I would rig a safety chain to keep the rear axle from raising more than a few inches and tilt my L4610 untill it started to lift a wheel and measure the angle. THEN I sure as he-- wouldn't get close to that tilt in actual use. Similarly I could do that front to rear with nose high and nose low variants.
Ballast, implements FEL, etc. would have much larger effect front to rear but would still introduce variability side to side as well.
Of course this would only determine the "NEVER EXCEED" angles for the particular configuration of implement, weights, driver weight, etc. In practice tractors are rough riding bouncy things that would most generally roll over if you ever got particularly close to the "STATIC" max tilt angle. The safe angle in use (dynamic situation) would always be less than the statically determined angle.
If anyone tries to make any measurements of this nature, be careful, very careful. If some of the engineering talent available on this board were put to this task they could probably make measurements of weight shift at tilts far below the tilt over angles and compute fairly closely what the ultimate angle would be without getting close to a roll over.
I think you could compute the CG by parking the left (or right) tires on scales at the feed store or wherever and jacking the opposite side of the tractor up. Record the measured weight increase vs the tilt angle up to maybe 10-15 degrees (something definitely safely away from a roll over). That's it. I think the rest is math. This should give the CG. If your tractor were square, same track width front and rear and probably same tire size you could compute max tilt angle before, in theory, the tractor would be ballanced on just two wheels. The data will likely plot as a curve so don't just measure two angles and interpolate or extrapolate.
The experiment should be repeated nose up and nose down as well. Since most of our tractors are not square, the math is probably a tad tougher on the side to side calcs. Anyway this should be easy enough to do and be pretty safe if using good shop practices when jacking it up. I bet someone could do this in 10-15 minutes if the feed store guy was alert and got the series of weight measurements recorded to coincide with your series of angles. The angle could be measured with a plumb bob and a protractor. That's all for now kids. Gee Mr. Wizard, what we gonna do next week?
Patrick