I hope I haven't come across as too negative, I do really appreciate the GRO making this data available to us. Just to be clear, all my comments as to what search options the GRO offer us, and may offer us, are purely gleaned from looking at the HTML and javascript on the page, so are speculative at best. I'm in no way associated with GRO other than as a user like everyone else here.
Regarding the mother's maiden name they do seem to be using a dash (ie no name) when the child was illegitimate. I'm comparing records with certificates I have. This seems slightly at odds with how it's done post-1911; some consistency would be good.
Regarding the quarter notation, like others, I keep doing a double take to convert it to Q1-Q4. I suspect it's the abbreviation to a single character that's the cause, as personally, I don't have the same issue when using FreeDMB and their 3 character abbreviations. I think I might change my script to 'fix' this.
As for my earlier comments on the phonetic name matching (metaphone), looking again at their HTML it appears they're not using exact matching on the metaphone code but using a level of fuzzy matching (ie 1 code character different). This might explain the performance issue and why the resulting matches are a bit looser than I/we expect. There's also a commented out option for even looser matches (ie 2 characters different), so it looks like they have may been experimenting with more options. An option for exact code matches would be nice.
Has anyone else wondered what this could mean for FreeDMB? As the GRO are building their own database with more detail than FreeBMD could (assuming they're never given access to the data the GRO have), could it affect the number of volunteers and donations FreeBMD can attract? I think it'd be a real shame if FreeBMD were 'relegated' to a secondary source through no fault of their own, considering all the work they've put in over the years.