Personally my recommendation would be to keep the jQuery backwards compatibility script until the next release branch is cut (which is in November, I think?). It does not block any further work (the upgrade already happened, the script just keeps some removed methods in existence), and for third parties who are working with the tarballs, switching from jQuery 1.8 to 1.11 without any help in transitioning will be harsh.

As far as multimedia issues are concerned: we agreed (as I understand it) to only work on TMH upgrade after the GWToolset load issues are resolved, since those actually block people from using the site, while depending on deprecated jQuery methods does not. If we want to revisit that decision, fixing TMH is probably not a huge amount of work - we could start on it now and be done by the original June 1 target date. If we stick to giving GWToolset higher priority - it's hard to estimate how long that will take. The optimistic scenario is that our current plans (rendering reference thumbnails on upload, using a reference thumbnail chain to speed up resizing and lighten the bandwidth load of scalers, use poolcounter to limit the number of expensive images that can be thumbnailed in parallel) are sufficient and turn out to be not too hard to implement, in which case... maybe a month? Gilles probably has a better grasp on the timeline. My optimistic estimates tend to be optimistic in the negative sense of the word :-)