Norman,
Personally and professionally, I'd say that if you're careful then removing the varnish is generally preferable. Primarily this is for four reasons:
1) The lacquer's composition is unknown, and undesirable impurities may have been added to it during mixture (i.e. the purity of the original solvent is unknown, in addition to potentially damaging impurities entering the compound from the crushing and liquefying stages).
2) The lacquer's integrity is unknown; an uneven coat may have worn away in certain spots while remaining thick in others, or may never have been properly applied at all. Liquid coatings such as shellac (and also nitrocellulose lacquers such as Frigilene) are always problematic in this regard since they tend to run, and I tend to think that a coating that gives a false sense of security is as bad as no coating at all.
3) The lacquer is likely not to be up to scratch vis-a-vis long-term conservation properties. Shellac undergoes hydrolysis into acids (aliphatic and alicyclic acids, no less - from what little chem I understand, the more unstable sorts of organic compounds), which are in turn likely to cause damage to the object if they aren't neutralised. Frigilene and other nitrocellulose lacquers, in accelerated aging tests, have tended to peel off, discolour, and become difficult to remove. I myself have observed the characteristic yellowing of frigilene lacquer on a sword in the RDG collection, the lacquer having been in place approximately 20 years by that time. There's
a rather good publication from the Getty Conservation Institute for anyone interested.
4) Frankly, that's a bloody ugly finish!
Let us know how you get on; there are a few objects in our collections that have been lacquered in their long (and long-suffering) lives, and I'd be interested to see the results of your work when planning any on these objects.
- Meredydd