It seems like the "anti-tank gun/rifle carrying cavalry" idea is the most sound proposition overall. We'd need a weapon - either an armour-piercing rifle, a large bore machine gun/autocannon, or a dedicated anti-armour gun based on a light artillery/infantry support gun - which could then be mounted on or carried by highly mobile cavalry-style beasts, which would also carry the gunners. Sort of a modernised
Zamburak. This is relatively economical, as the guns wouldn't be anything revolutionary and the beasts wouldn't need to be large or intelligent.
We could take that general concept and build on our existing success with the grenade projector though, and develop some sort of indirect-firing massed explosive barrage. The mobile cavalry-beasts would manoeuvre to be out of sight of the walkers, set up multiple salvos of cheap rocket-propelled grenades, and then saturate the target with explosive fire. It wouldn't need to be accurate provided we can pump out enough ammo to justify firing off dozens of the things at once, and since we're talking about light walkers a moderate explosive charge should be enough to cause spall to kill/harm the crew without necessarily needing to break through the armour. Indirect fire also means the beasts and crew need not be in direct line of sight of the walkers, which should reduce casualties and increase the likelihood of a successful ambush.
Given this, I'm going to vote as following:
[X] Captain Ernest Vroom
[X] Dr. Robert Montragard
[X] Lieutenant de vaisseau Matis Fyn
[X] Captain Dirck Lasoon
This is basically the consensus vote, but with Vandievoort switched out for Montragard. Lasoon is important for funding, and thus being able to do extra tests and iron out kinks in the design. Fyn has direct experience directly operating anti-walker weapons effectively, and so will be able to tell us more clearly what kind of thing does and does not work. Vroom is the developer of the Grenade Projector and I'd like to build on that design for this. And Montragard has experience developing new beasts for the French Cavalry, and so would be able to help develop a new, economical beast with maximum manoeuvrability for this project. Vandievoort is a good second choice for beast expert, as he's familiar with Belgium's current stock of beasts and we might be able to get away with just using an existing beast from stock, but since we have the money from Lasoon I'm inclined to be a bit more ambitious and try making a new beast specialised as a fast, manoeuvrable, all-terrain light weapons platform.