No I just don't agree.
All Ku Cable head end dishes and teleport dishes I've seen 6' or bigger using segmented construction have seams sealed. 1 mm is critical value for Ku. But your entitled to make subjective statments.
I have no intention of starting a flame war, but I must point out your statements above are as subjective, or more so, than anything B.J. or I wrote. If you can offer an objective analysis that demonstrates or proves typical dish seams can measurably degrade Ku performance, I would be very interested to see this. In the meantime perhaps I might propose a reasonable example to the contrary.
Let's make a gross estimate of the loss seams might have on a prime-focus, segmented dish, such as a 1.8 m Fortec or similar design. The effective collection surface area of such a dish is about 2.5 m^2. Such dishes have six petals, or six seams that are slightly less than the radius of the dish (my Fortec has a center reflector plate that is probably of no consequence as it is shadowed for the most part by the feed and scalar ring).
My Fortec's seams are no wider than say 5mm and have somewhat of a "V" cross-section. The maximum depth of the V is only a few mm. However let's assume the maximum depth is 1/4 wavelength (about 6mm for Ku-band) for analysis ease. This means a signal bouncing off the bottom of the "V" will be 180 degrees out-of-phase with the predominant dish reflection. A smaller depth will not achieve as much cancellation.
In reality the target signal impinging on a perfect "V" surface will be reflected away from the feed and will neither contribute, nor detract from the gain of the rest of the dish. The seam collection area is about 0.027 m^2, which reduces the collection area of our previously calculated 2.5 m^2. This causes a loss of 0.047 dB of gain with respect to a perfect surface. This is roughly equivalent to a 1.79 m perfect dish and is completely negligible.
To take this to the absurd conclusion, let's assume by magic the "V" reflects all of the signal straight to the feed, with only a path difference. What effect does this cause? Nothing. Because of the carefully chosen depth, any reflection caused at a depth "d" will be cancelled out by the reflection from the "6mm - d" depth. If the depth is not an odd multiple of 1/4 wavelength, the seam will actually provide some gain.
A couple of other remarks:
1. There is no 'critical value' for dish design such as the 1 mm you suggest. Any local departure from a true paraboloid will cause some phase shift. A 1 mm error is about 1/25 of a wavelength at Ku frequencies, and will result in about a 30 degree phase error. That's fairly small, but somewhat less or more will not change the performance by much.
2. As whatchel1 points out, the sealing of seams in commercial dishes improves the structural integrity of the dish. It also keeps out water, which happens to expand when frozen.
