I tend to think the overall aim is to increase rail's market share, then I think a few things are true:
Koln and Frankfurt in themselves don't represent a huge market.
Koln is a key destination for its onward connections.
Koln Hbf is problematic from a border security point of view, and I'm not convinced that Messe/Deutz is any better really - especially considering the very segregated running lines and platforms which would likely be disrupted by keeping 2 for international services.
Setting up another set of UK border controls elsewhere is likely to be financially and politically difficult.
As far as I can see, the logical way to increase rail's market share is therefore to improve connections to the existing London services; some of these are in the works, some from this thread:
Brussels to Switzerland
Lille to Barcelona
Amsterdam to Copenhagen
Brussels to Berlin
Brussels to Munich
Those would seem much more achievable and likely to achieve a reasonable amount of the possible modal share.
Yes - the current lack of variety in the major destinations served by onward connections from Brussels is a pain. And having direct (and fairly speedy!) Brussels-Switzerland day-time services, ditto Brussels-Berlin, would presumably have a market besides people changing to/from London; hence the "not enough people to fill a direct London-Switzerland train" doesn't apply.
That combination of truths re Cologne is a key problem.
Re Brussels-Munich - of course there already is the occasional Paris-Munich direct service (from Est, so not a very hassly interchange).
Overall, in the absence of London services going beyond their current terminals, the next best thing has to be through ticketing onto major long-distance services from Lille, Brussels and Paris (esp from Est). But that still leaves Italy - with Frejus closed for a year or so, a direct service from Lille (or Paris Est) to, say Milan, via Basel or Geneva would be a boon.
Decent connections, and through ticketing (with guarantees of no-fuss re-booking if a scheduled connection doesn't work out), are I think the key to increasing rail share of these journeys.
Though on reflection a Switzerland-Paris Est service would also work: even with a change at Strasbourg that route often comes up as fastest to London.
The annoying thing at the moment, returning from Switzerland to the UK, is having to choose between an unpleasant cross-Paris interchange, or a much easier one from Est, but with an inevitable change of trains en route (with the extra possibility of a delayed connection - as happened to me last time I did it that way).