Hmm, obviously we'll have to wait and see, but at least theoretically I think it _could_ be ok. My reasoning being that:
a) Oracle is at least a (successful) database company*. They understand the "product", they have the experience and technical resources. In the public mind, oracle=database. I'd be more concerned for SPARC, Solaris & Java (in that order). Oracle's not a hardware co, or an OS company (despite desire to have their own complete stack).
b) mySQL & Oracle's DB don't compete with each other. mySQL's popularity for dynamic websites isn't just down to it's price, but because (with the default MyISAM engine) it's optimised for read often, write seldom which works well in this context. Clustering support is nice to have, but I suspect it's mostly used for high-demand web apps. Oracle's db is in a different use space.
Oracle _could_ provide uniform tools, SQL dialect compatiblity, etc. for a clean upgrade path from mySQL (say as an intern department db) to Oracle when the need grows for some big iron.
If it goes pear shaped, well, the source code is there, Monty is still around live n kicking, and there's alway Postgres (which for those of us that program with a db abstraction layer should't be too much work). MySQL 5 is already installed all over the place, and does what it needs to do in 99% of cases.
As the poster says, Stay Calm & Carry On :-)
* Yes I know closed-source, and over-priced, but reasonably technically competent AFAIK.