If the initial data is gathered using Excel, I can upload it into a MySQL table using something called a cvs import to get you going. Once the initial data has been assembled, it can also be examined with a view to the best way to store it in MySQL. I appreciate that a flat file might, at first, seem to be the best way to proceed. And it's certainly the simplest approach. However, if this is truly intended to be a long term project with the capacity to handle a whole range of searches and queries, then a series of relational tables right from the start might be the best way to future proof the whole thing.
(For the less technically-minded, relational tables are simply a way of storing, for example, the ABOK reference name of a particular knot in one place, the information on its construction in another, it's potential uses in a 3rd table etc etc. Splitting the data up in this fashion greatly reduces the amount of work that the server needs to do to provide answers over a whole range of questions)
Once the data was imported into MySQL, attention would need to switch away from Excel and over to the central online system for all additions and amendments. Initially I could either write, or source, a basic front end system for all of the data management - ideally accessed by authorised individuals only using a username/password combination. Longer term, a more sophisticated admin interface could be gradually added. If this was tackled on a modular basis, it would limit any additional expenditure from the Guild's point of view and go some way to ensuring that the right tools were being built in the most appropriate way.
As soon as the data is online, work could also begin to allow users to run searches etc - either on the public site or within a restricted area.
Someone mentioned using Access. Frankly, if the intent is to get this online asap, I wouldn't bother. Access isn't well suited for online storage or usage - especially on the igkt.net server. Stick with Excel.
Ultimately, a full centralised, dynamic, online system is highly feasible. However, until I see some sample data, I can't really advise as to how it should eventually be organised other than to echo Knot Head's suggestions that a combination of PHP and MySql would be the way to go.