"is a MFJ 259B immune to the effect of showing inaccurate readings..."
No, it isn't. It's subject to the same inaccuracies any other device is. But it does give you readings that if understood can certainly make things easier. First, you have to know what it's telling you and why. That '259 can tell you things that a normal SWR meter "know's" absolutely nothing about.
"also , if a 259B shows different impedance on a antenna with different coax lengths is that normal ? or is there something wrong with the coax ?"
Yes, it can only tell you what it 'sees', and if that antenna isn't a 'perfect' load impedance (or close), it can only tell you what the combined impedance of the feed line and antenna are. It has no way of 'separating' those two things. That does not mean that there's something wrong with the coax. It simply means that the impedance of the antenna and the coax are not the same (or close).
One way of 'separating' the two thingys, antenna and feed line, is by using a known length of feed line that will show the same impedance at one end that' is present at the other end. An electrical 1/2 wave length feed line has that characteristic, so by using an electrical 1/2 wave length to measure antenna input impedance means that you can set/adjust the antenna from a point further away from that antenna (so your physical presence doesn't affect the readings you get). {One nice thing about that '259 is that it can tell you what an electrical 1/2 wave length is, you don't have to do any calculating. That's just a part of knowing how the @#$ thing works. The 'bad' part is that you have to know how all this stuff 'fits' together, it isn't intuitive. Gotta flex that muscle between them ears, you know? 
}
- 'Doc
Oops! Little late with a reply...
