The claim that this adds administrative overhead and increases implementation cost and decreases benefit in particular I think is incorrect.
Since most people will not participate in the "wasted-work" program (I need a pithy name for this scheme), the benefit per participant can be scaled up considerably from what UBI could provide on an individual basis.
Administratively I think it's probably close to a wash -- UBI has the administrative difficulty of preventing or detecting fraud (duplicate identities in particular), "wasted-work" has the administrative overhead of employing reliable "doormen" for the centers, and the real estate costs (and maintenance, etc.) of the centers, not to mention any administrative costs associated with attempts to fix the problems I highlighted around "commuting costs".
Both programs have the problem of setting the appropriate level of compensation and updating it to reflect changes in cost of living, but as I mentioned, "wasted-work" can be more sensitive to regional variations in the cost of living than UBI can (without a hugely increased administrative overhead).
Both of these are, obviously, lost in a presence-based minimum wage, which also abandons adds additional administrative overhead compared to UBI. AFAICT, this idea is strictly worse in terms of benefit and implementation cost.