From straight image quality terms, yes there will be a significant difference. Images from the 50-135 on the k1 will have heavily vingnetted and black corners or be limited to 15mp. The 28-105 will cover the whole frame, immediately giving it an advantage in the corners, and probably be close enough at f8 over the shared range with complete coverage. There is also no other Pentax designed option currently in production for the K-1 for that focal range, though the Tamron developed 24-70 is also excellent for that range on FF.
So in short, no, the 28-105 doesn't replace the 50-135 on aps-c. And the 50-135 isn't designed for the k-1, so yes the 28-105 is the better choice on FF.
But the 28-105 is not a replacement for getting the 70-200 to replace the 50-135 when moving from aps-c to the K-1, which is what I think you asked in your original post.
Despite the fact that the same lenses will work on both bodies, they are different formats. While you can go from FF to APS-C, you can't make the smaller format lenses cover the larger format anymore than you could expect a K lens to cover the sensor on a 645Z despite the fact that there are adapters to put a 645 lens on a K camera.
Originally posted by ShiftR Thanks for the replies. I am/was well aware that the lenses are designed for and aimed at completely different usages. I just wanted to get an idea as to whether the difference going between the two would be painfully obvious in strict IQ terms (not the "magic dust" rendering that the Lims and *s have). From the sounds of it, I may be okay until I can step up to the 70-200.