I know this is a crude example, but this is loosely equivalent to an election where voting booths are only placed in regions where significantly more people vote for party A, and then saying 'well, members of party B are allowed in those areas too'.
I agree that everyone is eligible but with the banner running across the top of the web version saying 'please check this thread and vote' and no equivalent within the mobile app, it will skew the responses in the sample to be more web browser users. I'm not saying it's done on purpose to skew the result, but it absolutely will. My uneducated guess would be that it's maybe an extra 10% so not making much difference, but given the topic of the vote, presumably there is no other information to cross this vote with to check if that guess is right.
This probably sounds like I'm desperately fighting for the app to remain in light of the rather black and white vote results today. While I do like the app, I'm just passionate about opinion research like this (which may be used to take a decision) is conducted fairly. Perhaps a DM out to members would be a better way to do it (for large total sample - which is presumably why the banner is used), or repeat the vote without the banner on the web version (will be lower total sample, but more likely to have natural fall out across platforms without the intrusiveness of DMs to everyone).