JimCasey, Thank you. I have more from the following month, but I start with tests on floats designed like those on hugho's red model, with the underside angled up from step to bow end.

I started designing floats based on the full size ones in the photo, on a Piper Cub at Jack Brown's seaplane base in Florida. Note that the bottom ahead of the step is parallel to the top for a bit, maybe a quarter of the total length.
Here is a bit of my next magazine article, dealing with that. Plus some systematic tests of float angle.

I tried a new design of floats on my latest PeTe design, the 46" span version with the OS 10. Until now, all my floats have had the planing bottom ahead of the step parallel to the top, but not this pair. This time I took someone's advice from a magazine article and angled the bottom up at 3 degrees. The photo will give the idea. They worked adequately into wind on choppy water, but when I tried them in calm weather the aeroplane behaved quite differently.
It was a real dog, refusing to take off and just squirreling about on the water like an ugly duckling. I tried changing the float angle, a wedge at the step, 30% nitro in the fuel. No good.
I decided to change the floats rather than the engine, since I recently discovered to my surprise that floatplanes can manage with remarkably little power. I went back to my normal float design and the model is directionally stable and will now run straight hands off. If the up trim is set it will take off by itself without pulling back on the stick as well, just like its three big brothers (56", 66", and 76" span). My advice is to make the floats with the bottom just ahead of the step parallel to the top and to the aircraft's datum line as that seems to work best.

Step position
The next obvious trial was to move the step too far forward, so the whole float mounting was unscrewed and re-attached with the step ahead of the CG. I put it in the water, opened the throttle and off it went. There were no problems taking off, or flying around, flying some aerobatics, or with circuits or touch and go landings. After burning off 10 oz of fuel I brought it in and cut the engine. As it glided to a stop in front of me the model leaned back until the tail slowly dipped into the water. All-right PeTe, point taken! The buoyancy is too far forward.
I took off the 45" floats and fitted a 36" pair which barely extended as far forward as the prop, with the step correctly placed. At rest in the water the floats were low in the water, betraying that they were meant to support 6 pounds, not 8. On calm water they threw up more spray initially as engine power pulled the nose down, until the model was up and planing. But on choppy water the model would sometimes surge forward and the prop would take a great bite out of the next approaching wave, creating a cloud of spray. Usually the engine recovered from the shock, but sometimes it would stop in protest. Not enough buoyancy forward.
Angle of Dangle
I was told, and I always tell others, mount the floats parallel to the datum line! But what if you don't?
On PeTe 10 the angle of the floats could conveniently be changed from level to five degrees either way in three increments. There are two ways of looking at this. Either you look at the model flying, with the nose of the floats angled either up or down. Or you sit it on its floats and say that the nose of the aeroplane is angled up or down relative to the floats. Since you sit it down before you try to fly, I'll refer to the second option.
I lengthened the rear float attachment to angle the nose of the aeroplane down by 1, 3 and then 5 degrees. In each case the model took off and flew fine though, especially on the last trial with the nose down at 5 degrees, the engine tended to pull the front of the floats down into the water creating excess spray. However, with a little speed the hydro-dynamic and aerodynamic forces lifted the float fronts, the model planed beautifully (directionally stable) and lifted off a treat. The handling was fine with just a little down trim needed. On landing the aircraft came in with the floats held nose up out of the way and when it hit the water and decelerated it pitched nose down onto the floats, which kills the lift and removes any thoughts it might have had about bouncing or porpoising.
Next the floats were angled the other way by up to 5 degrees, so that the aircraft sat nose up on the water. At 3 degrees the model would no longer track straight, but planed along chasing imaginary rabbits like a puppy in a field. It seemed that aerodynamic forces were lifting the tail to a level attitude, and that shifted the planing point from the step to the front of the planing surface, producing instability. However flight was achieved after a difficult take off.
While I was increasing the nose up sit to 5 degrees there was a sharp increase in the wind, producing a chop on the water. The exaggerated nose up sit of the model prevented take off, or rather, the plane leapt off the water far too soon and immediately cartwheeled back in. Several times. Eventually I gave up and hand launched. The model flew around well enough but it looked very odd and did not seem quite as stable in flight as normal. A touchdown on the water was attempted, and despite the nose down hang of the floats, was accomplished without incident.
To check that the refusal to take off was not attributable to the sudden rise in wind and rough water, the floats were returned to their usual angle and another take off was attempted. This time the model lifted off easily and cleanly from the water despite the rough surface.
So what have I learned?
1. If the floats are right excess power is not necessary, but power could compensate for other faults.
2. Put the step just aft of the CG, but it is not critical with excess power available.
3. Float size is not critical on calm water.
4. On rough water 10% of the float length ahead of the prop is a great help.
5. With the floats level on the bench, rig the model level (or slightly nose down).
6. A nose up sit (on the bench) produces planing instability and makes take off unexpectedly difficult.
7. Waterplane flying is always fun.