Its the same reason telescopes can't just "filter out" stuff like city lights and human radio sources.
Do you have any facts to show that the streaks raise the noise floor significantly (say more than 10%)?
I would guess from your lack of hard information that you are just suggesting it could potentially be a problem.
Edit: for a quick and dirty calculation, I estimated the pixels on a horizontal line in the image in the article, ignoring the galaxy, and the streak is not significant compared with the light from the stars. I presume because the satellites move fast across a slow time-lapse image capture, your hypothetical problem is not actually significant.