I can't speak for the example you give but in general the answer is: Use multiple accounts.
While we can increase the limits for some services well beyond what the defaults are, there will be side-effects. For example, having so many Lambda functions in a single account means that the concurrency limits will affect how many of those can be run at once and that (in turn) affects how you scale. Note that this is true of any account running many workloads - they can potentially get in each other's way as the limits are reached.
Again, limits can be increased but at some point it's better to be running those workloads in separate accounts so that they can scale independently of each other.
Account separation also means that you can delegate administration duties much easier, rather than trying to scope IAM policies to individual Lambda functions.
Is there overhead in running multiple accounts? Absolutely. But at the scale you're talking (thousands of functions) I would assume that you have a level of automation for those functions; and therefore you can automate the cross-account activities as well.
There is facility to request for increase in quotas under account in the console. See if you can request for what you want. If not, reach out to support and get their guidance.
How can we create thousands of Lambdas, beyond quota limits?asked 6 months ago
Best practices for hundreds/thousands of keys/usage plans?asked 3 years ago
limits with API gateway custom authorizer for number of requestsAccepted Answerasked 2 years ago
DynamoDB Tables Quota Increase Limitsasked 4 months ago
Access to S3- thousands of usersasked 2 months ago
How to create feed document XML to change the price of a ASIN for a seller?asked 3 months ago
Structure of Repositories for Resource creation and lambdasasked 5 months ago
Lambda Memory Quota Inconsistent in São Paulo regionasked 6 months ago
API rate limits for AWS SESasked 5 months ago
How to increase service quota limits?Accepted Answerasked 3 years ago