Just be careful about setting those mileposts up in advance. When you planned SW 2.0, you didn't know there'd be a 2.1 or 2.5 per se. Those revisions were essentially created by problems found that were not planned for.

So if your first ESB with completely-revised despecialization is 2.5, what's to say this version won't also have some unplanned fixes requiring future iterations? I know you're learning a lot as you go, and ESB might therefore go more according to plan, but still... seems safest to target something less than 2.5 just to give some wiggle room. Screw consistent versioning.