I think the main concern really is the "reliable support structure" aspect; I don't think being named "MySQL" is actually at all relevant to that assessment. What matters is that the software under active development is supported by established organizations providing support. This is actually probably the biggest concern, because the most obvious organization providing support is the one Oracle is buying (although other companies provide commercial support for MySQL). I think the fact that this organization is also the one that holds the trademarks isn't a major factor.
As far as development is concerned, I think that, if Oracle were to become too hard to work with, the community would take it over with approximately the efficiency that X.org took over from XFree86. If you've got a group of people who know each other as the development community, they're likely to keep working together. And, if Oracle owns the bulk of the copyrights and is unresponsive, there is no crown (in terms of a codebase that some branch can offer non-GPL licenses to) to fight over. It's less likely that everybody would try to take over the project from themselves than that everybody would try to sucker somebody else into doing the project administration.
Of course, there's still the question of whether third-party corporate support organizations would declare the Oracle branch dead and support the community branch instead. But I could easily see OpenLogic telling their customers, "Starting with the next release, MySQL as we support it has been renamed to OurSQL. It is the same software and the same contracts apply to it." And the corporate market would say, "Whatever. We still think our cell phones are Cingular. As long as our business relationships stay valid, it doesn't matter." And, of course, there's the issue that, if Oracle stops their MySQL development, the version that their support organization supports presumably also stops, and that support organization is probably the most significant one currently.