your western blot is more or less perfect...your bands are properly separated and not mixed
I think its not a major problem, until and unless you are able to detect your protein of interest by blotting.
Many reasons can be assigned to it like, if gel is not polymerized the bands tend to be diffused due to protein migration, If your loading buffer is too old you can have this problem.....
Try to polymerize the gel properly, I mean use it after 3-4 h after it is polymerized
I also faced this some times, but finally we have to use densitometric software to calibrate it...
I also agree with the researchers above that your western blot is very much perfect and also that the difference in the width is likely to be due to the difference in the buffer composition (if they are different for your samples)....similar things are seen in denaturing PAGE gels (containing UREA) used for separating DNA/RNA fragments ....if the well are not flushed properly....
the gel are prepared 1-3 days before the assay so this is probably not the issue. The samples are in the same buffers, so this is too may not the problem...
You are right that after all the amount is calculated in a densitometric software so the differences are diminished.
I can tell you clearly that problem is with that particular sample. I would like to say that do extensive dialysis of your protein sample which is wider in lanes may be for 16 h or o/n with repeated change of buffer or water . After that centrifuge your sample for 10-15 min at 4 C and use the supernatant for gel loading. Can you also tell what type of sample you are using i.e. extracellular or intracellular and also the source means plants or microbes? Believe me problem is not with your gel it is with sample. Please let me know after that about results. It will significantly increase the quality of your gel.