Yes, it is possible that the two traits could still be linked (assuming that they are not two aspects of the same trait in which case they would always be linked).
Two characteristics that are correlated by "accident", chance or serendipity will remain correlated unless there is active selection that breaks the correlation.
I am going to use the traditional daylily belief that evergreens are winter tender and dormants are winter hardy as an example.
Suppose that the evergreen species Hemerocallis aurantiaca var major (or more appropriately H. major) was tender and the original dormant species were hardy. That would be an original 100% correlation. It is likely that correlation would be less than 100% in the F1 between evergreen and dormant species. It might be even less or approximately the same in the F2. After the F2 generation the correlation would not change unless plants were actively selected to change the correlation.
Growing and hybridizing with evergreens in the north where winter conditions could damage or kill evergreens and not using those plants as parents would select for hardy evergreens and change the correlation. However, hybridizing in locations were selection could not occur against winter tender plants would not change the correlation. If natural selection acted negatively on winter-tender plants in other ways that affected their use as parents in locations that did not have severe winters then that could change the correlation.
Tets in general would have larger foliage than diploids. To be more accurate, the tetraploid version of a diploid plant would on average have wider foliage. However, if hybridizers considered tetraploid foliage that was too wide as "coarse" it could have been selected to be narrower and more similar to the width of diploid foliage.
